System for controling incinerator using artificial intelligence and operation method thereof

Abstract

A method of operating an incinerator control system (300), which includes a sensor unit (320), an operation control device (310), and a server (100) and controls an incinerator (410) using machine learning, includes: acquiring, by the operation control device, operation state information (831) for a predetermined analysis time from the sensor unit; acquiring, by the operation control device, operation setting information for the analysis time from a memory (220); acquiring (530) difference information between information at a current time and information at an immediately previous time based on the operation state information and the operation setting information; acquiring (540) trend information for the analysis time based on the operation state information and the operation setting information; and acquiring (550) predicted operation control information (840) by applying at least one of the operation state information, the operation setting information, the difference information, and the trend information to an operator simulation machine learning model (820) received from the server.

Description

Incinerator control system using artificial intelligence and method of operation of the system

[0001] The present disclosure relates to an incinerator control system and an operating method using artificial intelligence. More specifically, the system includes a method for optimizing combustion in the incinerator based on signals acquired from various sensors included in the incinerator.

[0002] Typically, incinerator design involves calculating the input waste conditions based on three representative criteria: high-quality, medium-quality, and low-quality. However, government policies and various practical circumstances sometimes lead to special situations where waste with different properties and calorific values ​​than those initially specified for the incinerator design is input. This leads to waste inputs outside the design scope and operating conditions that deviate from the design parameters.

[0003] This causes the waste fed into the incinerator to be incompletely combusted, and the combustion chamber temperature exceeds the designed value, which seriously affects the life and efficiency of the refractory, incinerator, and pollution prevention facilities. Looking at the specifics of the incinerator design, in the case of a general incinerator, a system is designed to supply secondary combustion air at approximately 20-40% of the total air supply to combust incomplete combustion products, tar, char, etc. contained in the unburned gas generated in the first combustion stage of the incinerator. In addition, the factors that determine the combustion performance of the incinerator are time, temperature, and turbulence, commonly referred to as the 3T. In order to destroy incomplete combustion products generated during the combustion process of the incinerator, mixing with air through strong turbulence at a certain temperature or higher is necessary. However, if the amount of combustion air is reduced, the velocity at the tip of the secondary combustion air nozzle is relatively reduced, resulting in insufficient destruction of incomplete combustion products within the incinerator. Meanwhile, since the detailed combustion situation inside an actual incinerator is very uneven and almost impossible to measure, a comprehensive incinerator control system that can represent the overall combustion situation using a heat balance program and the measured values ​​of the temperature inside the incinerator (incinerator outlet temperature, upper temperature of the drying stage, upper temperature of the post-combustion stage), the pressure inside the incinerator, and the oxygen concentration and carbon monoxide concentration of the combustion gas is required.

[0004] Conventional incinerator control systems strive to optimally control the incinerator by setting predefined control conditions. However, because these systems cannot predict the calorific value of the waste, they can only indicate whether the current combustion condition of the incinerator is normal, not whether it will operate normally in the future. Furthermore, if the surrounding environment, such as the waste, airflow, temperature, and humidity, does not meet the predefined control conditions, the system will not operate properly, requiring adaptive human intervention to determine whether to feed the waste. In other words, the automatic control capabilities of conventional incinerator control systems are inherently limited. Therefore, active research is being conducted on incinerator control systems that can operate normally even under conditions outside the designer's predefined conditions.

[0005] The incinerator control system of the present disclosure includes a sensor unit, an operation control device, and a server. The operation method of the incinerator control system uses machine learning, and includes a step in which the operation control device acquires operation status information for a predetermined analysis time from the sensor unit, a step in which the operation control device acquires operation setting information for the analysis time from a memory, a step in which the operation control device acquires difference information between a current time and a previous time based on the operation status information and the operation setting information, a step in which the operation status information and the operation setting information acquire trend information for the analysis time based on the operation status information and the operation setting information, and a step in which at least one of the operation status information, the operation setting information, the difference information, and the trend information is applied to a driver simulation machine learning model received from a server to acquire predicted operation control information.

[0006] The operating status information of the operating method of the incinerator control system of the present disclosure includes at least one of waste mass information, supply flow information, air flow information, pressure information, information about a storage tank, temperature information, humidity information, and exhaust gas information, and the waste mass information includes at least one of the input timing of waste currently being incinerated in the incinerator, the weight of the input waste, a predetermined calorific value of the input waste, a predetermined calorific curve of the input waste, and mass information of the waste input into the incinerator, and the supply flow information is related to the amount of at least one of urea water, ammonia water, ammonia gas, dilution water, slaked lime, magnesium hydroxide, an amount of water injected into the incinerator, deaerator feed water, cooling water feed water, economizer feed water, steam flow rate, or boiler feed water supplied to the incinerator, and the air flow information indicates the air flow rate of at least one of a pressure blower, an induced blower, or an exhaust gas recirculator, and the pressure information indicates the pressure inside the incinerator, the pressure inside the boiler, the pressure inside the cooling water pipe, and the boiler water pipe. pressure inside, pressure inside an ammonia tank, pressure inside a boiler drum, or steam pressure, and information about a storage tank is related to the degree of filling of at least one of a boiler feed water tank, a process water storage tank, a caustic soda feed tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a light oil storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, and temperature information is related to the inside of an incinerator, an incinerator outlet, an incinerator drying stage top, a boiler inlet, a boiler outlet, a selective non-catalytic reduction device inlet, a selective non-catalytic reduction device outlet, a semi-dry scrubber inlet, a semi-dry reactor outlet, a centrifugal dust collector (cyclone) inlet, a centrifugal dust collector (cyclone) outlet, a bag filter inlet, a bag filter outlet, a scrubber inlet, a scrubber outlet, a selective catalytic reduction device inlet, a selective catalytic reduction device outlet, Chimney exhaust, deaerator inlet, deaerator outlet, boiler feedwater temperature, steam temperature, economizer inlet, economizer outlet, feedwater tank, process water storage tank,At least one temperature of a caustic soda supply tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a washing water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, the atmosphere, a waste storage warehouse, a hopper section, a primary air supply, a secondary air supply, or an exhaust gas recirculation inlet or exhaust gas recirculation outlet, and an incinerator facility ambient temperature is related to the humidity information, and the humidity information is related to the inside of the incinerator, the incinerator outlet, the upper part of the incinerator drying stage, the boiler inlet, the boiler outlet, the selective non-catalytic reduction device inlet, the selective non-catalytic reduction device outlet, the semi-dry scrubber inlet, the semi-dry reactor outlet, the centrifugal dust collector (cyclone) inlet, the centrifugal dust collector (cyclone) outlet, the filter dust collector (bag filter) inlet, the filter dust collector (bag filter) outlet, the scrubber inlet, the scrubber outlet, the selective catalytic reduction device inlet, At least one of the humidity of a selective catalytic reduction device outlet, a stack exhaust, a water feed tank, a process water storage tank, a caustic soda supply tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, an atmosphere, a warehouse, a hopper section, a primary air supply, a secondary air supply, or an exhaust gas recirculation inlet or exhaust gas recirculation outlet, and an incinerator facility atmosphere, and information about the exhaust gas includes information about the amount of at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) emitted from the incinerator, and information about at least one of temperature, pressure, humidity, and flow rate.

[0007] The operation setting information of the operating method of the incinerator control system of the present disclosure includes at least one of blower air volume information, pusher information, blower damper opening rate information, chemical injection amount information for reducing air pollutants, chemical injection amount information for improving incineration efficiency, or waste input setting information, wherein the blower air volume information indicates the air volume of a pressure blower, an induced blower, or an exhaust gas recirculation blower, and further, the air volume indicates the output frequency of an electric motor inverter connected to the blower, the pusher information indicates the operation cycle or the number of operations of a pusher included in a ramp pusher, a drying stage stocker, a combustion stage stocker, or a post-combustion stage stocker, and the waste input setting information includes at least one of a waste input time point, an input waste weight, and a predetermined heating curve of the input waste.

[0008] The predicted operation control information of the operation method of the incinerator control system of the present disclosure includes one of the following: predicted information on whether waste is input, predicted blowing volume of an induced blower, predicted blowing volume of a pressurized blower, predicted blowing volume of an exhaust gas recirculation blower, predicted opening rate information of a blower damper, predicted chemical injection amount for reducing air pollutants, predicted chemical injection amount for improving incineration efficiency, or predicted pusher operation information, and the predicted information on whether waste is input indicates whether waste is input, the predicted blowing volume of the induced blower includes information on the predicted output frequency of an electric motor inverter connected to the induced blower, the predicted blowing volume of the pressurized blower includes information on the predicted output frequency of an electric motor inverter connected to the pressurized blower, the predicted blowing volume of the exhaust gas recirculation blower includes information on the predicted output frequency of an electric motor inverter connected to the exhaust gas recirculation blower, and the predicted opening rate information of the blower damper indicates whether waste is input, It indicates the predicted opening rate of a damper installed at least in one of the outlets of a pressurized blower and an exhaust gas recirculation blower, and the predicted chemical input amount information for reducing air pollutants includes at least one of the type, timing, and amount of the input chemical, and the predicted chemical input amount information for improving incineration efficiency includes at least one of the type, timing, and amount of the input chemical, and the predicted pusher operation information includes at least one of the cycle in which the pusher pushes the waste and whether the pusher is operated (on / off).

[0009] The method of operating the incinerator control system of the present disclosure further includes a step of controlling the incinerator by an operation control device based on predicted operation control information, and a step of displaying at least one of predicted waste input information, predicted blower volume information of an induced blower, predicted blower volume information of a pressure blower, predicted exhaust gas recirculation blower volume information, predicted blower damper opening rate information, or predicted pusher operation information.

[0010] The predicted operation control information of the operation method of the incinerator control system of the present disclosure indicates predicted information on whether or not waste is input, and the step of controlling the incinerator by the operation control device based on the predicted operation control information includes a step of the operation control device inputting waste when the predicted information on whether or not waste is input indicates that waste is input, and a step of the operation control device not inputting waste when the predicted information on whether or not waste is input indicates that waste is not input.

[0011] The step of obtaining differential information of the operation method of the incinerator control system of the present disclosure includes the step of obtaining a plurality of pieces of learning operation status information and a plurality of pieces of operation setting information at a basic sampling cycle, the step of obtaining operation status information of the current time, operation setting information of the current time, operation status information of the previous time, and operation setting information of the previous time according to a system cycle based on the plurality of pieces of learning operation status information and the plurality of pieces of operation setting information, and the step of obtaining differential information by differentiating one of the operation status information of the current time and the operation setting information of the current time from one of the operation status information of the previous time and the operation setting information of the previous time, wherein the difference between the current time and the previous time is a system cycle.

[0012] The trend information of the operation method of the incinerator control system of the present disclosure includes change amount information and change direction information of one of the operation status information and the operation setting information for each system cycle during the analysis time.

[0013] The step of obtaining trend information of the operating method of the incinerator control system of the present disclosure includes the steps of obtaining a plurality of pieces of learning operation status information and a plurality of pieces of operation setting information at a basic sampling cycle, obtaining operation status information at an n-th time, operation setting information at an n-th time, operation status information at an n-1-th time, and operation setting information at an n-1-th time according to a system cycle based on the plurality of pieces of learning operation status information and the plurality of pieces of operation setting information, obtaining a plurality of pieces of difference information by differentiating one of the operation status information at an n-1-th time and the operation setting information at an n-1-th time from one of the operation status information at an n-th time and the operation setting information at an n-1-th time, and obtaining the plurality of pieces of difference information as trend information, wherein the difference between the n-th time and the n-1-th time is the system cycle, and the magnitude of the plurality of pieces of difference information corresponds to change amount information and the sign of the plurality of pieces of difference information corresponds to change direction information.

[0014] The operating method of the incinerator control system of the present disclosure includes a step in which a server obtains a plurality of pieces of learning operation status information, a plurality of pieces of learning operation setting information, and a plurality of pieces of operation control information from a pre-collected learning database, a step in which the server obtains a plurality of pieces of past differential information and a plurality of pieces of past trend information based on the plurality of pieces of learning operation status information and the plurality of pieces of learning operation setting information, a step in which the server generates a driver simulation machine learning model by machine learning a correlation (causal relationship) of the plurality of pieces of operating control information with respect to at least one of the plurality of pieces of learning operation status information, the plurality of pieces of learning operation setting information, the plurality of pieces of past differential information, and the plurality of pieces of past trend information, and a step in which the driver simulation machine learning model is transmitted to an operation control device.

[0015] The step of obtaining a plurality of pieces of learning operation status information, a plurality of pieces of learning operation setting information, and a plurality of pieces of operation control information of the operating method of the incinerator control system of the present disclosure includes a step of obtaining candidate learning operation status information, candidate learning operation setting information, and candidate operation control information, a step of obtaining operation status information corresponding to the candidate learning operation status information, the candidate learning operation setting information, and the candidate operation control information after a predetermined time, a step of obtaining proficiency information based on the operation status information after a predetermined time, and a step of determining the candidate operation control information as one of the plurality of pieces of operation control information when the proficiency information is equal to or greater than a predetermined threshold compensation value.

[0016] The incinerator control system of the present disclosure includes a sensor unit, an operation control device, and a server, and an operation method of the incinerator control system that controls the incinerator using machine learning includes a step in which the operation control device acquires operation status information for a predetermined analysis time from the sensor unit, a step in which the operation control device acquires operation setting information for the analysis time from a memory, a step in which the operation control device acquires difference information between a current time and a previous time based on the operation status information and the operation setting information, a step in which the operation status information and the operation setting information acquire trend information for the analysis time based on the operation status information and the operation setting information, a step in which at least one of the operation status information, the operation setting information, the difference information, and the trend information is applied to an operation status prediction machine learning model received from a server to acquire predicted operation status information, and a step in which the predicted operation status information is operation status information at a future time after a predetermined prediction time from the current time.

[0017] The operating status information of the operating method of the incinerator control system of the present disclosure includes at least one of waste mass information, supply flow information, blower flow information, pressure information, information about a storage tank, temperature information, humidity information, and exhaust gas information, and the waste mass information includes at least one of the input time of waste currently being incinerated in the incinerator, the weight of the input waste, a predetermined calorific value of the input waste, a predetermined calorific curve of the input waste, and mass information of the waste input into the incinerator, and the supply flow information is related to the amount of at least one of urea water, ammonia water, ammonia gas, dilution water, slaked lime, magnesium hydroxide, an amount of water injected into the incinerator, deaerator feed water, cooling water feed water, economizer feed water, steam flow rate, or boiler feed water supplied to the incinerator, and the blower flow information represents the blower flow rate of at least one of a pressure blower, an induced blower, or an exhaust gas recirculator, and the pressure information represents the pressure inside the incinerator, the pressure inside the boiler, the pressure inside the cooling water pipe, and the pressure inside the boiler water pipe. Pressure, pressure in an ammonia tank, pressure in a boiler drum, or steam pressure, and information about a storage tank is related to the degree of filling of at least one of a boiler feed water tank, a process water storage tank, a caustic soda feed tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a light oil storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, and temperature information is related to the inside of the incinerator, the outlet of the incinerator drying stage, the boiler inlet, the boiler outlet, the selective non-catalytic reduction device inlet, the selective non-catalytic reduction device outlet, the semi-dry scrubber inlet, the semi-dry reactor outlet, the centrifugal dust collector (cyclone) inlet, the centrifugal dust collector (cyclone) outlet, the bag filter inlet, the bag filter outlet, the scrubber inlet, the scrubber outlet, the selective catalytic reduction device inlet, the selective catalytic reduction device outlet, Chimney exhaust, deaerator inlet, deaerator outlet, boiler feedwater temperature, steam temperature, economizer inlet, economizer outlet, water feed tank,At least one temperature of a process water storage tank, a caustic soda supply tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, the atmosphere, a waste storage warehouse, a hopper section, a primary air supply, a secondary air supply, or an exhaust gas recirculation inlet or exhaust gas recirculation outlet, and an incinerator facility atmosphere temperature is related to the humidity information, and the humidity information is related to the inside of the incinerator, the incinerator outlet, the upper part of the incinerator drying stage, the boiler inlet, the boiler outlet, the selective non-catalytic reduction device inlet, the selective non-catalytic reduction device outlet, the semi-dry scrubber inlet, the semi-dry reactor outlet, the centrifugal dust collector (cyclone) inlet, the centrifugal dust collector (cyclone) outlet, the filter dust collector (bag filter) inlet, the filter dust collector (bag filter) outlet, the scrubber inlet, the scrubber outlet, and the selective catalytic reduction device. At least one of the humidity of an inlet, a selective catalytic reduction device outlet, a stack exhaust, a feed water tank, a process water storage tank, a caustic soda supply tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, an atmosphere, a warehouse, a hopper section, a primary air supply, a secondary air supply, or an exhaust gas recirculation inlet or exhaust gas recirculation outlet, and an incinerator facility atmosphere, and information about the exhaust gas includes information about the amount of at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) emitted from the incinerator, and information about at least one of temperature, pressure, humidity, and flow rate.

[0018] The operation setting information of the operating method of the incinerator control system of the present disclosure includes at least one of blower air volume information, pusher information, blower damper opening rate information, chemical injection amount information for reducing air pollutants, chemical injection amount information for improving incineration efficiency, or waste input setting information, wherein the blower air volume information indicates the air volume of a pressure blower, an induced blower, or an exhaust gas recirculation blower, and further, the air volume indicates the output frequency of an electric motor inverter connected to the blower, the pusher information indicates the operation cycle or the number of operations of a pusher included in a ramp pusher, a drying stage stocker, a combustion stage stocker, or a post-combustion stage stocker, and the waste input setting information includes at least one of a waste input time point, an input waste weight, and a predetermined heating curve of the input waste.

[0019] The step of outputting predicted operating status information of the operating method of the incinerator control system of the present disclosure includes the step of determining whether the predicted operating status information is included in predetermined critical environment information, and the step of automatically generating predicted operating control information when the predicted operating status information is not included in the critical environment information.

[0020] The predicted operating status information of the operating method of the incinerator control system of the present disclosure includes at least one of predicted temperature information, predicted oxygen amount, and predicted pressure information, and the critical environment information includes at least one of critical temperature range information, critical oxygen amount range, and critical pressure range information.

[0021] The step of obtaining differential information of the operation method of the incinerator control system of the present disclosure includes the step of obtaining a plurality of pieces of learning operation status information and a plurality of pieces of operation setting information at a basic sampling cycle, the step of obtaining operation status information of the current time, operation setting information of the current time, operation status information of the previous time, and operation setting information of the previous time according to a system cycle based on the plurality of pieces of learning operation status information and the plurality of pieces of operation setting information, and the step of obtaining differential information by differentiating one of the operation status information of the current time and the operation setting information of the current time from one of the operation status information of the previous time and the operation setting information of the previous time, wherein the difference between the current time and the previous time is a system cycle.

[0022] The system cycle of the operating method of the incinerator control system of the present disclosure is longer than the basic sampling cycle, and the step of acquiring the operating status information of the current time, the operating setting information of the current time, the operating status information of the previous time, and the operating setting information of the previous time includes the step of acquiring the operating status information of the current time, the operating setting information of the current time, the operating status information of the previous time, and the operating setting information of the previous time by averaging a plurality of pieces of learned operating status information and a plurality of pieces of operating setting information for each system cycle.

[0023] The trend information of the operation method of the incinerator control system of the present disclosure includes change amount information and change direction information of one of the operation status information and the operation setting information for each system cycle during the analysis time.

[0024] The step of obtaining trend information of the operating method of the incinerator control system of the present disclosure includes the steps of obtaining a plurality of pieces of learning operation status information and a plurality of pieces of operation setting information at a basic sampling cycle, obtaining operation status information at an n-th time, operation setting information at an n-th time, operation status information at an n-1-th time, and operation setting information at an n-1-th time according to a system cycle based on the plurality of pieces of learning operation status information and the plurality of pieces of operation setting information, obtaining a plurality of pieces of difference information by differentiating one of the operation status information at an n-1-th time and the operation setting information at an n-1-th time from one of the operation status information at an n-th time and the operation setting information at an n-1-th time; and obtaining the plurality of pieces of difference information as trend information, wherein the difference between the n-th time and the n-1-th time is the system cycle, the magnitude of the plurality of pieces of difference information corresponds to change amount information, and the sign of the plurality of pieces of difference information corresponds to change direction information.

[0025] The operating method of the incinerator control system of the present disclosure includes a step in which a server acquires a plurality of pieces of learning operation status information, a plurality of pieces of learning operation setting information, and a plurality of pieces of operation status information after a prediction time from a previously collected learning database, a step in which the server acquires a plurality of pieces of past differential information and a plurality of pieces of past trend information based on the plurality of pieces of learning operation status information and the plurality of pieces of learning operation setting information, a step in which the server generates an operation status prediction machine learning model by machine learning a causal relationship (correlation) of operation status information after a plurality of prediction times for at least one of the plurality of pieces of learning operation status information, the plurality of pieces of learning operation setting information, the plurality of pieces of past differential information, and the plurality of pieces of past trend information, and a step in which the operation status prediction machine learning model is transmitted to an operation control device, wherein the operation status information after a plurality of prediction times is information after a prediction time after acquiring the plurality of pieces of learning operation status information.

[0026] The incinerator control system of the present disclosure includes a sensor unit, an operation control device, and a server, and an operation method of the incinerator control system that controls the incinerator using machine learning includes a step in which the operation control device acquires operation status information for a predetermined analysis time from the sensor unit, a step in which the operation control device acquires operation setting information for the analysis time from a memory, a step in which the operation control device acquires difference information between a current time and a previous time based on the operation status information and the operation setting information, a step in which the operation control device acquires trend information for the analysis time based on the operation status information and the operation setting information, a step in which the operation control device determines at least one of the operation status information, the operation setting information, the difference information, and the trend information as current status information, and a step in which the operation control device applies the current status information to an incinerator control reinforcement learning model received from a server to acquire predicted operation control information that provides a maximum reward.

[0027] The operating method of the incinerator control system of the present disclosure further includes a step of generating an incinerator control reinforcement learning model by performing reinforcement learning on a server so that a reward by a reward function (r) is maximized, and a step of transmitting the incinerator control reinforcement learning model to an operation control device, wherein the reward function (r) is r = r_temp + r_NOX + r_CO + r_HCL + r_SOX + r_Dust + r_waste, where r_temp is a reward value according to learning temperature information, r_NOX is a reward value according to the amount of learning nitrogen oxide, r_CO is a reward value according to the amount of learning carbon monoxide, r_HCL is a reward value according to the amount of learning hydrogen chloride, r_SOX is a reward value according to the amount of learning sulfur oxide, r_Dust is a reward value according to the amount of learning dust, and r_waste is a reward value according to the amount of waste input per hour, and learning status information includes learning temperature information, learning nitrogen oxide amount, learning carbon monoxide amount, learning It includes at least one of the amount of hydrogen chloride, the amount of learning sulfur oxide, the amount of learning dust, and the amount of waste input per hour.

[0028] The step of generating an incinerator control reinforcement learning model of the operating method of the incinerator control system of the present disclosure includes a step in which a server acquires k-th learning operation control information from among a plurality of candidate operation control information available in k-th learning state information based on policy information, a step in which the server reflects the k-th learning operation control information into the k-th learning state information to acquire k+1-th learning state information, a step in which the server applies the k+1-th learning state information to a reward function to determine a k-th sub-reward of the k-th learning operation control information for the k-th learning state information, a step in which the server determines the k-th reward based on the k-th sub-reward and the k+1-th sub-reward, a step in which the server changes the policy information so that the k-th reward is maximized, and a step in which the server generates the incinerator control reinforcement learning model based on the policy information.

[0029] The step of determining the kth sub-compensation of the operating method of the incinerator control system of the present disclosure comprises: a step in which the server determines r_temp as a first compensation value when learning temperature information included in the k+1th learning state information is equal to or greater than a first threshold temperature and less than a second threshold temperature; a step in which the server determines r_temp as a second compensation value when learning temperature information included in the k+1th learning state information is equal to or greater than a third threshold temperature and less than a first threshold temperature, or equal to or greater than a second threshold temperature and less than a fourth threshold temperature; a step in which the server determines r_temp as a third compensation value when learning temperature information included in the k+1th learning state information is equal to or greater than a third threshold temperature and less than a fourth threshold temperature; and a step in which the server determines r_temp as a fourth compensation value when learning temperature information included in the kth learning state information is equal to or greater than a fifth threshold temperature and or greater than a sixth threshold temperature, and learning temperature information included in the k+1th learning state information is equal to or greater than a seventh threshold temperature and less than a third threshold temperature, or equal to or greater than a fourth threshold temperature and less than an eighth threshold temperature. Including, the first compensation value, the second compensation value, and the fourth compensation value are positive, the third compensation value is negative, the second compensation value is smaller than the absolute values ​​of the first compensation value and the third compensation value, and the relationship of the fifth critical temperature < the seventh critical temperature < the third critical temperature < the first critical temperature < the second critical temperature < the fourth critical temperature < the eighth critical temperature < the sixth critical temperature is satisfied.

[0030] The step of determining the kth sub-compensation of the operating method of the incinerator control system of the present disclosure includes the step of the server determining r_NOX as a fifth compensation value when the amount of learned nitrogen oxide included in the k+1th learning state information is less than a predetermined amount of reference nitrogen oxide, the step of the server determining r_NOX as a sixth compensation value when the amount of learned nitrogen oxide included in the k+1th learning state information is greater than or equal to the amount of reference nitrogen oxide and less than a predetermined amount of allowable nitrogen oxide, and the step of the server determining r_NOX as a seventh compensation value when the amount of learned nitrogen oxide included in the k+1th learning state information is greater than or equal to the amount of allowable nitrogen oxide, wherein the fifth compensation value is positive, the sixth compensation value and the seventh compensation value are negative, the absolute values ​​of the fifth compensation value and the sixth compensation value are less than the absolute value of the seventh compensation value, and the amount of reference nitrogen oxide is less than the amount of allowable nitrogen oxide.

[0031] The step of determining the kth sub-compensation of the operating method of the incinerator control system of the present disclosure includes the step of the server determining r_CO as an eighth compensation value when the amount of learned carbon monoxide included in the k+1th learning state information is less than a predetermined reference carbon monoxide amount, the step of the server determining r_CO as a ninth compensation value when the amount of learned carbon monoxide included in the k+1th learning state information is greater than or equal to the reference carbon monoxide amount and less than a predetermined allowable carbon monoxide amount, and the step of the server determining r_CO as a tenth compensation value when the amount of learned carbon monoxide included in the k+1th learning state information is greater than or equal to the allowable carbon monoxide amount, wherein the eighth compensation value is a positive number, the ninth compensation value and the tenth compensation value are negative numbers, the absolute values ​​of the eighth compensation value and the ninth compensation value are less than the absolute value of the tenth compensation value, and the amount of the reference carbon monoxide is less than the amount of the allowable carbon monoxide.

[0032] The step of determining the kth sub-compensation of the operating method of the incinerator control system of the present disclosure includes the step of the server determining r_HCL as an eleventh compensation value when the amount of learning hydrogen chloride included in the k+1th learning state information is less than a predetermined amount of reference hydrogen chloride, the step of the server determining r_HCL as a twelfth compensation value when the amount of learning hydrogen chloride included in the k+1th learning state information is greater than or equal to the amount of reference hydrogen chloride and less than a predetermined amount of allowable hydrogen chloride, and the step of the server determining r_HCL as a thirteenth compensation value when the amount of learning hydrogen chloride included in the k+1th learning state information is greater than or equal to the amount of allowable hydrogen chloride, wherein the eleventh compensation value is positive, the twelfth compensation value and the thirteenth compensation value are negative, the absolute values ​​of the eleventh compensation value and the twelfth compensation value are less than the absolute value of the thirteenth compensation value, and the amount of reference hydrogen chloride is less than the amount of allowable hydrogen chloride.

[0033] The step of determining the kth sub-compensation of the operating method of the incinerator control system of the present disclosure includes the step of the server determining r_SOX as a 14th compensation value when the amount of learned sulfur oxide included in the k+1th learning state information is less than a predetermined amount of reference sulfur oxide, the step of the server determining r_SOX as a 15th compensation value when the amount of learned sulfur oxide included in the k+1th learning state information is greater than or equal to the amount of reference sulfur oxide and less than a predetermined amount of allowable sulfur oxide, and the step of the server determining r_SOX as a 16th compensation value when the amount of learned sulfur oxide included in the k+1th learning state information is greater than or equal to the amount of allowable sulfur oxide, wherein the 14th compensation value is positive, the 15th compensation value and the 16th compensation value are negative, the absolute values ​​of the 14th compensation value and the 15th compensation value are less than the absolute value of the 16th compensation value, and the amount of reference sulfur oxide is less than the amount of allowable sulfur oxide.

[0034] The step of determining the kth sub-compensation of the operating method of the incinerator control system of the present disclosure includes the step of the server determining r_Dust as a 17th compensation value when the amount of learning dust included in the k+1th learning state information is less than a predetermined amount of reference dust, the step of the server determining r_Dust as an 18th compensation value when the amount of learning dust included in the k+1th learning state information is greater than or equal to the amount of reference dust and less than a predetermined amount of allowable dust, and the step of the server determining r_Dust as a 19th compensation value when the amount of learning dust included in the k+1th learning state information is greater than or equal to the amount of allowable dust, wherein the 17th compensation value is positive, the 18th compensation value and the 19th compensation value are negative, the absolute values ​​of the 17th compensation value and the 18th compensation value are less than the absolute value of the 19th compensation value, and the amount of reference dust is less than the amount of allowable dust.

[0035] The step of determining the kth sub-compensation of the operating method of the incinerator control system of the present disclosure determines r_waste by the following equation, r_waste = - a * |amount of reference waste - amount of waste input per unit time|, where a is a predetermined real number, the amount of reference waste is a predetermined value, and the amount of waste input per unit time is at least one of the mass and weight of waste input per unit time into the incinerator.

[0036] The operating status information of the operating method of the incinerator control system of the present disclosure includes at least one of waste mass information, supply flow information, air flow information, pressure information, information about a storage tank, temperature information, humidity information, and exhaust gas information, and the waste mass information includes at least one of the input timing of waste currently being incinerated in the incinerator, the weight of the input waste, a predetermined calorific value of the input waste, a predetermined calorific curve of the input waste, and mass information of the waste input into the incinerator, and the supply flow information is related to the amount of at least one of urea water, ammonia water, ammonia gas, dilution water, slaked lime, magnesium hydroxide, an amount of water injected into the incinerator, deaerator feed water, cooling water feed water, economizer feed water, steam flow rate, or boiler feed water supplied to the incinerator, and the air flow information indicates the air flow rate of at least one of a pressure blower, an induced blower, or an exhaust gas recirculator, and the pressure information indicates the pressure inside the incinerator, the pressure inside the boiler, the pressure inside the cooling water pipe, and the boiler water pipe. pressure inside, pressure inside an ammonia tank, pressure inside a boiler drum, or steam pressure, and information about a storage tank is related to the degree of filling of at least one of a boiler feed water tank, a process water storage tank, a caustic soda feed tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a light oil storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, and temperature information is related to the inside of an incinerator, an incinerator outlet, an incinerator drying stage top, a boiler inlet, a boiler outlet, a selective non-catalytic reduction device inlet, a selective non-catalytic reduction device outlet, a semi-dry scrubber inlet, a semi-dry reactor outlet, a centrifugal dust collector (cyclone) inlet, a centrifugal dust collector (cyclone) outlet, a bag filter inlet, a bag filter outlet, a scrubber inlet, a scrubber outlet, a selective catalytic reduction device inlet, a selective catalytic reduction device outlet, Chimney exhaust, deaerator inlet, deaerator outlet, boiler feedwater temperature, steam temperature, economizer inlet, economizer outlet, feedwater tank, process water storage tank,At least one temperature of a caustic soda supply tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a washing water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, the atmosphere, a waste storage warehouse, a hopper section, a primary air supply, a secondary air supply, or an exhaust gas recirculation inlet or exhaust gas recirculation outlet, and an incinerator facility ambient temperature is related to the humidity information, and the humidity information is related to the inside of the incinerator, the incinerator outlet, the upper part of the incinerator drying stage, the boiler inlet, the boiler outlet, the selective non-catalytic reduction device inlet, the selective non-catalytic reduction device outlet, the semi-dry scrubber inlet, the semi-dry reactor outlet, the centrifugal dust collector (cyclone) inlet, the centrifugal dust collector (cyclone) outlet, the filter dust collector (bag filter) inlet, the filter dust collector (bag filter) outlet, the scrubber inlet, the scrubber outlet, the selective catalytic reduction device inlet, At least one of the humidity of a selective catalytic reduction device outlet, a stack exhaust, a water feed tank, a process water storage tank, a caustic soda supply tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, an atmosphere, a warehouse, a hopper section, a primary air supply, a secondary air supply, or an exhaust gas recirculation inlet or exhaust gas recirculation outlet, and an incinerator facility atmosphere, and information about the exhaust gas includes information about the amount of at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) emitted from the incinerator, and information about at least one of temperature, pressure, humidity, and flow rate.

[0037] In addition, a program for implementing the operating method of the incinerator control system of the present disclosure as described above can be recorded on a computer-readable recording medium.

[0038] The present disclosure provides an incinerator control system using artificial intelligence and an operating method of the system, which optimizes combustion in the incinerator, minimizes the emission of pollutants measured by the incinerator TMS (Tele-Monitoring System) by maintaining appropriate conditions within the incinerator, minimizes the emission of incinerator ash, maximizes steam production, and preserves / optimizes the incinerator facilities.

[0039] FIG. 1 is a block diagram of a server according to one embodiment of the present disclosure.

[0040] FIG. 2 is a diagram illustrating a server according to one embodiment of the present disclosure.

[0041] FIG. 3 is a drawing for explaining the configuration of an incinerator control system according to one embodiment of the present disclosure.

[0042] Figure 4 illustrates an incinerator according to one embodiment of the present disclosure.

[0043] FIG. 5 is a flowchart illustrating the operation of an incinerator control system according to one embodiment of the present disclosure.

[0044] FIG. 6 is a drawing for explaining an incinerator control system according to one embodiment of the present disclosure.

[0045] FIG. 7 is a flowchart illustrating the operation of an incinerator control system according to one embodiment of the present disclosure.

[0046] FIG. 8 is a diagram illustrating a driver simulation machine learning model according to one embodiment of the present disclosure.

[0047] FIG. 9 is a flowchart illustrating the operation of an incinerator control system according to another embodiment of the present disclosure.

[0048] FIG. 10 is a diagram for explaining a driving state prediction machine learning model according to one embodiment of the present disclosure.

[0049] FIG. 11 is a flowchart illustrating an operation method of an incinerator control system according to another embodiment of the present disclosure.

[0050] FIG. 12 may be a drawing for explaining the operation of an incinerator control system according to one embodiment of the present disclosure.

[0051] FIG. 13 is a drawing for explaining the operation of an incinerator control system according to one embodiment of the present disclosure.

[0052] The advantages and features of the disclosed embodiments, and the methods for achieving them, will become clearer with reference to the embodiments described below, along with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below and may be implemented in various different forms. These embodiments are provided solely to ensure the completeness of the disclosure and to fully inform those skilled in the art of the present disclosure of the scope of the invention.

[0053] The terms used in this specification will be briefly explained, and the disclosed embodiments will be described in detail.

[0054] The terms used in this specification have been selected from widely used, current terms, taking into account the functions of the present disclosure. However, these terms may vary depending on the intentions of engineers working in the relevant fields, precedents, the emergence of new technologies, etc. Furthermore, in certain cases, terms may be arbitrarily selected by the applicant, and in such cases, their meanings will be described in detail in the relevant description of the invention. Therefore, the terms used in this disclosure should not be defined simply as names, but rather based on the meanings of the terms and the overall content of the present disclosure.

[0055] In this specification, singular expressions include plural expressions unless the context clearly indicates that they are singular. In addition, plural expressions include singular expressions unless the context clearly indicates that they are plural.

[0056] When a part of a specification is said to "include" a component, this does not mean that it excludes other components, but rather that it may include other components, unless otherwise stated.

[0057] Also, the term "part" used in the specification means a software or hardware component, and the "part" performs certain functions. However, the "part" is not limited to software or hardware. The "part" may be configured to reside on an addressable storage medium and may be configured to execute one or more processors. Thus, by way of example, the "part" includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functionality provided within the components and "parts" may be combined into a smaller number of components and "parts" or further separated into additional components and "parts."

[0058] According to one embodiment of the present disclosure, a "unit" may be implemented as a processor and a memory. The term "processor" should be broadly interpreted to include a general-purpose processor, a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a controller, a microcontroller, a state machine, and the like. In some circumstances, a "processor" may also refer to an application-specific integrated circuit (ASIC), a programmable logic device (PLD), a programmable logic controller (PLC), a field programmable gate array (FPGA), and the like. The term "processor" may also refer to a combination of processing devices, such as, for example, a combination of a DSP and a microprocessor, a combination of a plurality of microprocessors, a combination of one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0059] The term "memory" should be interpreted broadly to include any electronic component capable of storing electronic information. The term memory may also refer to various types of processor-readable media, such as random access memory (RAM), read-only memory (ROM), non-volatile random access memory (NVRAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable PROM (EEPROM), flash memory, magnetic or optical data storage, registers, etc. A memory is said to be in electronic communication with a processor if the processor can read information from and / or write information to the memory. Memory integrated in a processor is in electronic communication with the processor.

[0060] Below, with reference to the attached drawings, a detailed description of the embodiments is provided so that those skilled in the art can easily implement the present disclosure. Furthermore, in order to clearly illustrate the present disclosure, portions irrelevant to the description are omitted from the drawings.

[0061] FIG. 1 is a block diagram of a server (100) according to one embodiment of the present disclosure.

[0062] Referring to FIG. 1, a server (100) according to one embodiment may include at least one of a data learning unit (110) or a data recognition unit (120). The server (100) as described above may include a processor and a memory. FIG. 1 illustrates that the server (100) includes the data learning unit (110) or the data recognition unit (120), but is not limited thereto. The driving control device may include at least one of the data learning unit (110) or the data recognition unit (120). The server (100) may include some of the components included in the data learning unit (110) or the data recognition unit (120), and the driving control device (310) may include the remaining components included in the data learning unit (110) or the data recognition unit (120). The server (100) and the driving control device (310) can implement a data learning unit (110) or a data recognition unit (120) by exchanging information wired or wirelessly.

[0063] The data learning unit (110) can learn a machine learning model for performing a target task using a data set. The data learning unit (110) can receive label information related to the data set and the target task. The data learning unit (110) can perform machine learning on the relationship between the data set and the label information to obtain a machine learning model. The machine learning model obtained by the data learning unit (110) can be a model for generating label information using the data set.

[0064] The data recognition unit (120) may receive and store the machine learning model of the data learning unit (110). The data recognition unit (120) may apply the machine learning model to input data and output label information. In addition, the data recognition unit (120) may use the input data, label information, and results output by the machine learning model to update the machine learning model.

[0065] At least one of the data learning unit (110) and the data recognition unit (120) may be manufactured in the form of at least one hardware chip and mounted on an electronic device. For example, at least one of the data learning unit (110) and the data recognition unit (120) may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or may be manufactured as a part of an existing general-purpose processor (e.g., CPU or application processor) or a graphics-only processor (e.g., GPU) and mounted on various electronic devices as described above.

[0066] In addition, the data learning unit (110) and the data recognition unit (120) may be mounted on separate electronic devices, respectively. For example, one of the data learning unit (110) and the data recognition unit (120) may be included in the electronic device, and the other may be included in the server. In addition, the data learning unit (110) and the data recognition unit (120) may provide the machine learning model information constructed by the data learning unit (110) to the data recognition unit (120) via wired or wireless communication, and data input to the data recognition unit (120) may be provided to the data learning unit (110) as additional learning data.

[0067] Meanwhile, at least one of the data learning unit (110) and the data recognition unit (120) may be implemented as a software module. When at least one of the data learning unit (110) and the data recognition unit (120) is implemented as a software module (or a program module including instructions), the software module may be stored in a memory or a non-transitory computer readable medium that can be read by a computer. In addition, in this case, at least one software module may be provided by an operating system (OS) or by a predetermined application. Alternatively, some of at least one software module may be provided by an operating system (OS), and the remaining some may be provided by a predetermined application.

[0068] A data learning unit (110) according to one embodiment of the present disclosure may include a data acquisition unit (111), a preprocessing unit (112), a learning data selection unit (113), a model learning unit (114), and a model evaluation unit (115).

[0069] The data acquisition unit (111) can acquire data required for machine learning. Since a large amount of data is required for learning, the data acquisition unit (111) can receive a data set containing multiple pieces of data.

[0070] Label information can be assigned to each of the plurality of data. The label information can be information describing each of the plurality of data. The label information can be information that the target task seeks to derive. The label information can be obtained from user input, memory, or the results of a machine learning model. For example, if the target task is to predict the operating status information after a predicted time based on the operating status information and past operating settings information of an incinerator, the plurality of data used for machine learning would be the operating status information and past operating settings information of the incinerator, and the label information would be the operating status information after the predicted time. The predicted time can be a value of 1 second or more and 4 hours or less. The predicted time can be a value of 2 minutes or more and 1 hour or less. The predicted time can be a value of 5 minutes or more and 1 hour or less. The label information can be a sensor value obtained from a sensor unit or information directly entered by the user. In addition, for example, if the target task is to predict operation control information based on at least one of the operation status information of the incinerator or the past operation setting information, the plurality of data used for machine learning will be the operation status information and the past operation setting information of the incinerator, and the label information will be the operation control information after the predicted time. The label information may be control information directly input by the user. However, it is not limited thereto, and the label information may be a value obtained by rule-based operation. For example, the label information may be current operation control information obtained rule-based based on at least one of the operation status information of the incinerator or the operation setting information.

[0071] Additionally, in the case of a task that calculates the calorific value according to the type and amount of waste, the type and amount of waste will be multiple data, and the calorific value can be obtained based on a formula (rule base) that calculates the calorific value according to the type and amount of waste.

[0072] The preprocessing unit (112) can preprocess the acquired data so that the received data can be used for machine learning. The preprocessing unit (112) can process the acquired data set into a preset format so that the model learning unit (114) described later can use it. The preprocessing unit (112) of the server (100) or the driving control device can obtain differential information or trend information based on the driving status information or driving setting information. The preprocessing unit (112) of the server (100) or the driving control device can further use the obtained differential information or trend information to obtain driving control information or driving status information.

[0073] The learning data selection unit (113) can select data required for learning from among the preprocessed data. The selected data can be provided to the model learning unit (114). The learning data selection unit (113) can select data required for learning from among the preprocessed data according to preset criteria. In addition, the learning data selection unit (113) can also select data according to preset criteria through learning by the model learning unit (114) described below.

[0074] The model learning unit (114) can learn criteria regarding what label information to output based on the data set. Furthermore, the model learning unit (114) can perform machine learning using the data set and label information for the data set as learning data. Furthermore, the model learning unit (114) can additionally perform machine learning using a previously acquired machine learning model. In this case, the previously acquired machine learning model may be a pre-built model. For example, the machine learning model may be a pre-built model that receives basic learning data as input.

[0075] A machine learning model can be constructed by considering the application field of the learning model, the purpose of learning, or the computer performance of the device. The machine learning model can be, for example, a Boosting series ML algorithm, a tree-based algorithm, or a model based on a neural network. For example, models such as AdaBoost, GBM (Gradient Boosting Machine), XGBoost (Extra gradient boost), LightBoost, decision tree, random forest, deep neural network (DNN), recurrent neural network (RNN), long short-term memory models (LSTM), BRDNN (Bidirectional Recurrent Deep Neural Network), and convolutional neural networks (CNN) can be used as machine learning models, but are not limited thereto.

[0076] According to various embodiments, when there are multiple pre-built machine learning models, the model learning unit (114) may determine a machine learning model with a high correlation between the input learning data and the basic learning data as the machine learning model to be learned. In this case, the basic learning data may be pre-classified by data type, and the machine learning model may be pre-classified by data type. For example, the basic learning data may be pre-classified based on various criteria such as the location where the learning data was generated, the time when the learning data was generated, the size of the learning data, the creator of the learning data, the type of object in the learning data, etc.

[0077] Additionally, the model learning unit (114) can learn a machine learning model using a learning algorithm, such as error back-propagation or gradient descent, for example.

[0078] In addition, the model learning unit (114) can learn a machine learning model, for example, through supervised learning using learning data as input values. In addition, the model learning unit (114) can acquire a machine learning model, for example, through unsupervised learning that discovers a standard for a target task by independently learning the types of data required for the target task without any special guidance. In addition, the model learning unit (114) can learn a machine learning model, for example, through reinforcement learning that utilizes feedback on whether the results of the target task according to learning are correct.

[0079] Additionally, once the machine learning model is trained, the model training unit (114) can store the trained machine learning model. In this case, the model training unit (114) can store the trained machine learning model in the memory of an electronic device including a data recognition unit (120). Alternatively, the model training unit (114) can store the trained machine learning model in the memory of a server connected to the electronic device via a wired or wireless network.

[0080] The memory in which the trained machine learning model is stored may also store commands or data related to at least one other component of the electronic device, for example. The memory may also store software and / or programs. Programs may include, for example, a kernel, middleware, an application programming interface (API), and / or an application program (or "application").

[0081] The model evaluation unit (115) inputs evaluation data into the machine learning model, and if the result output from the evaluation data does not satisfy a predetermined standard, it can cause the model learning unit (114) to relearn. In this case, the evaluation data may be preset data for evaluating the machine learning model. In addition, the model evaluation unit (115) relearns the model based on data for a specific period of time at a predetermined period of time, and if the accuracy of the relearned model is higher than that of the existing model, the relearned model can replace the existing model. Here, the predetermined period of time may be 1 second or more and 4 years or less. In addition, the predetermined period of time may be 1 day or more and 1 month or less. In addition, the specific period of time may be 1 day or more and 1 month or less. The predetermined period of time and the specific period of time may be the same or different.

[0082] For example, the model evaluation unit (115) may evaluate that a predetermined criterion is not satisfied if the number or ratio of evaluation data with incorrect recognition results among the results of the trained machine learning model for the evaluation data exceeds a preset threshold. For example, if the predetermined criterion is defined as a ratio of 2%, if the trained machine learning model outputs incorrect recognition results for more than 20 evaluation data out of a total of 1,000 evaluation data, the model evaluation unit (115) may evaluate that the trained machine learning model is not suitable.

[0083] Meanwhile, if there are multiple trained machine learning models, the model evaluation unit (115) can evaluate whether each trained machine learning model satisfies a predetermined criterion and determine the model that satisfies the predetermined criterion as the final machine learning model. In this case, if there are multiple models that satisfy the predetermined criterion, the model evaluation unit (115) can determine one or a predetermined number of models, which are preset in order of highest evaluation score, as the final machine learning model.

[0084] Meanwhile, at least one of the data acquisition unit (111), the preprocessing unit (112), the learning data selection unit (113), the model learning unit (114), and the model evaluation unit (115) within the data learning unit (110) may be manufactured in the form of at least one hardware chip and mounted on an electronic device. For example, at least one of the data acquisition unit (111), the preprocessing unit (112), the learning data selection unit (113), the model learning unit (114), and the model evaluation unit (115) may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or may be manufactured as a part of an existing general-purpose processor (e.g., CPU or application processor) or a graphics-only processor (e.g., GPU) and mounted on the various electronic devices described above.

[0085] In addition, the data acquisition unit (111), the preprocessing unit (112), the learning data selection unit (113), the model learning unit (114), and the model evaluation unit (115) may be mounted on a single electronic device, or may be mounted on separate electronic devices. For example, some of the data acquisition unit (111), the preprocessing unit (112), the learning data selection unit (113), the model learning unit (114), and the model evaluation unit (115) may be included in the electronic device, and the remaining parts may be included in the server.

[0086] In addition, at least one of the data acquisition unit (111), the preprocessing unit (112), the learning data selection unit (113), the model learning unit (114), and the model evaluation unit (115) may be implemented as a software module. When at least one of the data acquisition unit (111), the preprocessing unit (112), the learning data selection unit (113), the model learning unit (114), and the model evaluation unit (115) is implemented as a software module (or a program module including instructions), the software module may be stored in a non-transitory computer readable medium that can be read by a computer. In addition, in this case, at least one software module may be provided by an operating system (OS) or may be provided by a predetermined application. Alternatively, some of at least one software module may be provided by an operating system (OS), and the remaining some may be provided by a predetermined application.

[0087] A data recognition unit (120) according to one embodiment of the present disclosure may include a data acquisition unit (121), a preprocessing unit (122), a recognition data selection unit (123), a recognition result provision unit (124), and a model update unit (125).

[0088] The data acquisition unit (121) can receive input data. The preprocessing unit (122) can preprocess the acquired input data so that the acquired input data can be used in the recognition data selection unit (123) or the recognition result provision unit (124).

[0089] The recognition data selection unit (123) can select required data from among the preprocessed data. The selected data can be provided to the recognition result provision unit (124). The recognition data selection unit (123) can select some or all of the preprocessed data according to preset criteria. In addition, the recognition data selection unit (123) can also select data according to preset criteria through learning by the model learning unit (114).

[0090] The recognition result providing unit (124) can obtain result data by applying the selected data to a machine learning model. The machine learning model may be a machine learning model generated by the model learning unit (114). The recognition result providing unit (124) can output result data. For example, the recognition result providing unit (124) can receive the operation status information and operation setting information of the incinerator and output the operation status information or predicted operation control information after the predicted time as result data. In addition, the recognition result providing unit (124) can receive the operation status information and operation setting information of the incinerator and output the operation status information after the predicted time as result data. In addition, the recognition result providing unit (124) can also generate the operation control information predicted based on rules based on the current operation status information and the operation status information after the predicted time.

[0091] The model update unit (125) can update the machine learning model based on an evaluation of the recognition results provided by the recognition result provider (124). For example, the model update unit (125) can provide the recognition results provided by the recognition result provider (124) to the model learning unit (114), thereby causing the model learning unit (114) to update the machine learning model.

[0092] Meanwhile, at least one of the data acquisition unit (121), the preprocessing unit (122), the recognition data selection unit (123), the recognition result provision unit (124), and the model update unit (125) within the data recognition unit (120) may be manufactured in the form of at least one hardware chip and mounted on an electronic device. For example, at least one of the data acquisition unit (121), the preprocessing unit (122), the recognition data selection unit (123), the recognition result provision unit (124), and the model update unit (125) may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or may be manufactured as a part of an existing general-purpose processor (e.g., CPU or application processor) or a graphics-only processor (e.g., GPU) and mounted on the various electronic devices described above.

[0093] In addition, the data acquisition unit (121), the preprocessing unit (122), the recognition data selection unit (123), the recognition result provision unit (124), and the model update unit (125) may be mounted on a single electronic device, or may be mounted on separate electronic devices. For example, some of the data acquisition unit (121), the preprocessing unit (122), the recognition data selection unit (123), the recognition result provision unit (124), and the model update unit (125) may be included in the electronic device, and the remaining parts may be included in the server.

[0094] In addition, at least one of the data acquisition unit (121), the preprocessing unit (122), the recognition data selection unit (123), the recognition result provision unit (124), and the model update unit (125) may be implemented as a software module. When at least one of the data acquisition unit (121), the preprocessing unit (122), the recognition data selection unit (123), the recognition result provision unit (124), and the model update unit (125) is implemented as a software module (or a program module including instructions), the software module may be stored in a non-transitory computer readable medium that can be read by a computer. In addition, in this case, at least one software module may be provided by an operating system (OS) or by a predetermined application. Alternatively, some of at least one software module may be provided by an operating system (OS), and the remaining some may be provided by a predetermined application.

[0095] Below, the method and device for receiving and processing learning data by the data acquisition unit (111), preprocessing unit (112), and learning data selection unit (113) of the data learning unit (110) are described in more detail.

[0096] FIG. 2 is a diagram illustrating a server according to one embodiment of the present disclosure.

[0097] The server (100) may include a processor (210) and a memory (220). The processor (210) may execute instructions stored in the memory (220).

[0098] As described above, the server (100) or the driving control device may include a data learning unit (110) or a data recognition unit (120). The server (100) may include at least one of a data acquisition unit (111), a preprocessing unit (112), a learning data selection unit (113), a model learning unit (114), a model evaluation unit (115), a data acquisition unit (121), a preprocessing unit (122), a recognition data selection unit (123), a recognition result provision unit (124), and a model update unit (125) within the data recognition unit (120). The data learning unit (110) or the data recognition unit (120) may be implemented by a processor (210) and a memory (220).

[0099] Although FIGS. 1 and 2 describe the server (100), the present invention is not limited thereto. The driving control device (310) to be described below may also include the same configuration as the server (100). That is, the driving control device (310) may also include a data learning unit (110) or a data recognition unit (120). In addition, the driving control device (310) may include a processor (210) and a memory (220). The processor (210) may execute commands stored in the memory (220). The driving control device (310) may include at least one of a data acquisition unit (111), a preprocessing unit (112), a learning data selection unit (113), a model learning unit (114), a model evaluation unit (115), a data acquisition unit (121), a preprocessing unit (122), a recognition data selection unit (123), a recognition result provision unit (124), and a model update unit (125) within the data recognition unit (120). The driving control device (310) can generate a machine learning model by exchanging data with the server (100) or derive result data using the machine learning model.

[0100] FIG. 3 is a drawing for explaining the configuration of an incinerator control system according to one embodiment of the present disclosure.

[0101] The incinerator control system (300) can control the incinerator using a machine learning model. The incinerator control system (300) can include an operation control device (310), a sensor unit (320), and a server (100). In the present disclosure, the description that the incinerator control system (300) performs operations may mean that operations are performed by at least one of the server (100) and the operation control device (310) included in the incinerator control system (300).

[0102] The operation control device (310) may include a PC, a smart phone, a PDA, a laptop, a desktop, a programmable logic controller (PLC), or a wearable device. The operation control device (310) may be a device located in the incinerator and may be a device used to collect data from the sensor unit (320) and control the incinerator. The operation control device (310) may transmit a sensor control signal to the sensor unit (320). In addition, the operation control device (310) may receive and store sensor signals from the sensor units (320) installed at various locations in the incinerator. In addition, the operation control device (310) may control operation setting information such as whether or not to input waste into the incinerator, the blower's air volume, the opening rate of the blower damper, the amount of chemicals injected to reduce air pollutants, the amount of chemicals injected to improve incineration efficiency, and the operation cycle or number of operations of the pusher (grate) using user input or artificial intelligence. Additionally, the airflow, which is one of the driving setting information, can be controlled by adjusting the output frequency of the electric motor inverter connected to the blower.

[0103] The sensor unit (320) may include various sensors installed in the incinerator. The sensor unit (320) may obtain operating status information. The operating status information may include at least one of waste mass information, supply flow rate information, blower flow rate information, pressure information, information about a storage tank, temperature information, humidity information, and information about exhaust gas. That is, the sensor unit (320) may include at least one of a mass sensor, a supply flow rate measurement sensor, a blower flow rate measurement sensor, a pressure sensor, a level measurement sensor, a temperature sensor, a humidity sensor, an exhaust gas detection sensor, or an exhaust gas concentration measurement sensor.

[0104] Additionally, the incinerator control system may include a server (100). The server (100) may communicate with the operation control device (310) via wired or wireless communication. The server (100) may be located remotely from the operation control device (310), but is not limited thereto, and may be located in close proximity. The server (100) may receive data from the operation control device (310) and store or process the data. The server (100) may receive at least one of operation status information, operation setting information, or operation control information from the operation control device (310).

[0105] The server (100) can process data from the driving control device (310) to obtain result information. The result information may be a machine learning model obtained based on at least one of driving status information, driving setting information, and driving control information. The server (100) can transmit the result information to the driving control device (310). The driving control device (310) can obtain predicted driving status information or predicted driving control information using the received machine learning model. According to another embodiment of the present disclosure, the server (100) may not transmit the machine learning model to the driving control device (310). The driving control device (310) can transmit driving status information and driving setting information to the server (100) to be input into the machine learning model. The server (100) can apply the driving status information and driving setting information to the stored machine learning model to obtain driving control information or driving status information after a predicted time as result information. Here, the predicted time may be 1 second or more and 4 hours or less. In addition, the predicted time may be a value of 5 minutes or more and 1 hour or less. The server (100) may transmit the result information to the operation control device (310). The operation control device (310) may output the received predicted operation status information or predicted operation control information. The user may control the incinerator based on the predicted operation status information or predicted operation control information. In addition, the operation control device (310) may automatically control the blower, pusher, and whether or not to input waste into the incinerator based on the predicted operation control information. More specifically, the operation control device (310) may automatically control operation setting information such as whether or not to input waste into the incinerator, the blower air volume, the opening rate of the blower damper, the amount of chemicals injected to reduce air pollutants, the amount of chemicals injected to improve incineration efficiency, and the operation cycle or number of operations of the pusher (grate) based on the predicted operation control information. Additionally, the airflow, which is one of the driving setting information, can be controlled by adjusting the output frequency of the electric motor inverter connected to the blower.

[0106] In FIG. 3, the incinerator control system (300) is disclosed as including an operation control device (310) and a server (100). However, the configuration of the incinerator control system (300) is not limited to FIG. 3. The incinerator control system (300) may include only the operation control device (310). That is, operations performed in the server (100) may be performed by the operation control device (310). In addition, the incinerator control system (300) may include only the server (100). That is, operations performed in the operation control device (310) may also be performed by the server (100).

[0107] The operation method of the incinerator control system (300) is described in more detail below.

[0108] Figure 4 illustrates an incinerator according to one embodiment of the present disclosure.

[0109] The incinerator (410) may be coupled with a polluted gas treatment facility (420), and the polluted gas treatment facility (420) may remove environmental pollutants from the exhaust gas. Referring to Fig. 4, the incinerator (410) and the polluted gas treatment facility (420) may be functionally distinct, but may not be physically distinct.

[0110] The polluted gas treatment facility (420) is composed of a selective non-catalytic reduction (SNCR) (421) for removing acid gases and air pollutants from combustion air, a centrifugal dust collection facility (cyclone) (422), a semi-dry scrubber (semi-dry reactor) (423), a wet scrubber (scrubber) (424), a filter dust collection facility (bag filter) (425), a selective catalytic reduction (SCR) (426), an induced blower (427), and a chimney (428) for discharging combustion gas into the atmosphere. The induced blower included in the incinerator can induce combustion gas to be discharged into the atmosphere through the chimney.

[0111] The driving control device (310) can measure the amount and calorific value of incinerator material fed into the incinerator (410), mass data measured by the meter of the garbage feeding crane, and control the moving speed and stopping speed of the pusher (grate). In addition, the operation control device (310) can control the operation of the incinerator by considering the temperature detected by the temperature sensor included in the combustion chamber (431), the pressure detected by the pressure sensor, the air volume of the primary and secondary pressure blowers, the air volume of the exhaust gas recirculation blower, the air volume of the induced blower, and the variables detected by the sensors included in the boiler (435), and the hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) detected by the selective non-catalytic reduction device (SNCR) (421), centrifugal dust collector (cyclone) (422), semi-dry scrubber (semi-dry reactor) (423), wet scrubber (scrubber) (424), filtration dust collector (bag filter) (425), selective catalytic reduction device (SCR) (426), and chimney (428), and can also selectively perform such control.

[0112] In this way, the operation control device (310) of the incinerator (410) can control the incinerator by detecting the steam generation amount of the boiler and the incinerator outlet temperature and analyzing them. At this time, the boiler steam generation amount, steam temperature, steam pressure, properties and calorific value of the incinerator, the air volume (air volume) of the primary and secondary pressure blowers, the air volume of the exhaust gas recirculation device, the air volume of the induced blower, the speed of the feeder and grate (pusher), and at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), and oxygen (O2) may be major variable values ​​for the operation control device (310) to control the incinerator.

[0113] Looking at the primary combustion air supply system of the incinerator, the primary combustion air nozzles (457, 458, 459) installed at the bottom of the primary combustion chamber of the incinerator (410) can be controlled so that the tip velocity of the outlet nozzle is maintained above a predetermined value. The main dampers (452, 453, 454, 455, 456) of the primary pressure blower (451) serve to supply the required amount of air within the incinerator (410), and when the main dampers (452, 453, 454, 455, 456) are 100% opened, the total amount of combustion air can be supplied into the incinerator (410) through each of the primary combustion air nozzles (457, 458, 459).

[0114] Looking at the secondary combustion air supply system of the incinerator, the secondary combustion air nozzle (443) installed at the bottom of the secondary combustion chamber of the incinerator (410) can be controlled so that the end velocity of the outlet nozzle is maintained above a predetermined value. The main damper (442) of the secondary pressure blower (441) serves to supply the required amount of air within the incinerator (410), and when the main damper (442) is opened 100%, the total amount of combustion air can be supplied into the incinerator (410) through each secondary combustion air nozzle (443).

[0115] Looking at the exhaust gas recirculation supply system of the incinerator, the primary combustion air nozzle (457, 458, 459) installed at the bottom of the primary combustion chamber of the incinerator (410), the secondary combustion air nozzle (443) installed at the bottom of the secondary combustion chamber, and the tip flow rate of the outlet nozzle can be controlled to be maintained above a predetermined value. The main damper (462, 463, 464) of the exhaust gas recirculation blower (461) of the incinerator serves to supply the required amount of air inside the incinerator (410), and when the main damper (462, 463, 464) is 100% opened, the recirculation air can be supplied into the incinerator (410) through each of the primary combustion air nozzles (457, 458, 459) and the secondary combustion air nozzle (443).

[0116] In order to ensure sufficient contact with the combustion gas of the incinerator (410) during operation of the incinerator, the discharge velocity of each primary combustion air nozzle (457, 458, 459) must be maintained at a specific velocity. At this time, the main damper (452, 453, 454, 455, 456) of the primary pressure blower (451) can be maintained in an open state.

[0117] In order to ensure sufficient contact with the combustion gas of the incinerator (410) during operation of the incinerator, the discharge velocity of each secondary combustion air nozzle (443) must maintain a jet stream. At this time, the main damper (442) of the secondary pressure blower (441) can be maintained in an open state.

[0118] In order to ensure sufficient contact with the combustion gas of the incinerator (410) during operation of the incinerator, the discharge velocity of each of the primary combustion air nozzles (457, 458, 459) and the secondary combustion air nozzle (443) must maintain a jet stream. At this time, the main damper (462, 463, 464) of the exhaust gas recirculation blower (461) can be maintained in an open state.

[0119] In the drying stage (432) of the incinerator, combustion processes such as moisture evaporation, volatile matter combustion in the combustion stage (433), and fixed carbon combustion in the post-combustion stage (434) can take place.

[0120] FIG. 5 is a flowchart illustrating the operation of an incinerator control system according to one embodiment of the present disclosure.

[0121] The incinerator control system (300) can recommend operation control information or predict operation status information using a machine learning model or reinforcement learning model. Accordingly, a user can determine the operation direction of the incinerator based on the recommended operation control information or predicted operation status information. The following describes the operation method of the incinerator control system (300).

[0122] The driving control device (310) may perform a step (510) of acquiring driving status information from the sensor unit (320) for a predetermined analysis time. The predetermined analysis time may be a time for collecting data required for the machine learning model to generate result information. The driving control device (310) may acquire driving status information in real time from the sensor unit (320). The sensor unit (320) may transmit a sensor signal to the driving control device (310) at a basic sampling period that is shorter than the predetermined analysis time. In addition, the driving control device (310) may resample the sensor signal acquired at the basic sampling period to a system period. The system period may be longer than the basic sampling period and shorter than the analysis time. The predetermined analysis time may have a value of 1 second or more and 4 hours or less. In addition, the predetermined analysis time may have a value of 1 minute or more and 1 hour or less. Additionally, the driving control device (310) transmits the sensor signal acquired at the basic sampling cycle to the server (100), and the server (100) may resample the received sensor signal. Although resampling is described below based on the driving control device (310), it should be interpreted that it can be performed by the server (100).

[0123] The operating status information may be information including at least one of temperature, humidity, pressure, liquid level, amount of material, or concentration of material that indicates the current status of the incinerator.

[0124] The driving status information may include at least one of waste mass information, supply flow information, blower flow information, pressure information, storage tank liquid level information, temperature information, humidity information, and exhaust gas information.

[0125] The mass information of the waste may include at least one of the input time of the waste currently being incinerated in the incinerator, the weight of the input waste, the predetermined calorific value of the input waste, and the predetermined calorific curve of the input waste. Furthermore, the mass information of the waste may include the mass information of the waste input into the incinerator. Furthermore, the mass information of the waste may include the mass information of the waste that has been input into the hopper by a crane but has not yet been input into the incinerator and is scheduled to be input in the future. Furthermore, the mass information of the waste may include the accumulated amount input by hour or the accumulated amount input since a reset time. The reset time may be, for example, 0:00 every day, but is not limited thereto.

[0126] The supply flow rate information may be related to the amount of at least one of urea water, ammonia water, ammonia gas, dilution water, slaked lime, magnesium hydroxide, incinerator water injection amount, deaerator feed water, cooling water feed water, economizer feed water, steam flow rate, or boiler feed water supplied to the incinerator. The supply flow rate information may have a unit related to the flow rate of the fluid. The supply flow rate information may be measured by a flow rate sensor included in the sensor unit (320).

[0127] The blower flow rate information may indicate the blower flow rate of at least one of the pressure blower (441, 451), the induced blower, or the exhaust gas recirculator. The blower flow rate information may include the blower flow rate of air or oxygen for burning waste in the incinerator. The pressure blower may include a primary pressure blower and a secondary pressure blower (441). The pressure blower may be a blower for blowing air into the incinerator. The induced blower may be a blower for discharging the exhaust from the incinerator to a chimney. The exhaust gas recirculator may be a blower for resupplying air from the incinerator to the incinerator. The blower flow rate information may be measured by a flow sensor included in the sensor unit (320).

[0128] The pressure information may be related to at least one of the pressure within the incinerator, the pressure within the boiler, the pressure within the cooling water pipe, the pressure within the boiler water pipe, the pressure within the ammonia tank, the pressure within the boiler drum, or the steam pressure. The pressure information may be measured by a pressure sensor included in the sensor unit (320).

[0129] Information about the storage tank may include information about the liquid level of the storage tank. The information about the storage tank may relate to the degree of filling of at least one of a boiler feed water tank, a process water storage tank, a caustic soda supply tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank. The information about the storage tank may be measured by a material detection sensor included in the sensor unit (320) or a means for detecting the level of a material within the tank.

[0130] Temperature information is provided at the inside of the incinerator, the incinerator outlet, the top of the incinerator drying stage, the boiler inlet, the boiler outlet, the selective non-catalytic reduction device inlet, the selective non-catalytic reduction device outlet, the semi-dry scrubber inlet, the semi-dry reactor outlet, the centrifugal dust collector (cyclone) inlet, the centrifugal dust collector (cyclone) outlet, the bag filter inlet, the bag filter outlet, the scrubber inlet, the scrubber outlet, the selective catalytic reduction device inlet, the selective catalytic reduction device outlet, the stack discharge, the deaerator inlet, the deaerator outlet, the boiler feedwater temperature, the steam temperature, the economizer inlet, the economizer outlet, the service water feed tank, the process water storage tank, the caustic soda supply tank, the slaked lime slurry storage tank, the urea water storage tank, the ammonia storage tank, the diesel storage tank, the scrubber storage tank, the wastewater storage tank, the process water storage tank, the city water storage tank, or the magnesium hydroxide tank, the atmosphere, and the waste storage. The temperature may be related to at least one of the warehouse, hopper, primary air supply, secondary air supply, exhaust gas recirculation inlet or exhaust gas recirculation outlet, and incinerator facility ambient temperature. The temperature information may be measured by a temperature sensor included in the sensor unit (320).

[0131] Humidity information is provided at the inside of the incinerator, the incinerator outlet, the top of the incinerator drying stage, the boiler inlet, the boiler outlet, the selective non-catalytic reduction device inlet, the selective non-catalytic reduction device outlet, the semi-dry scrubber inlet, the semi-dry reaction tower outlet, the centrifugal dust collector (cyclone) inlet, the centrifugal dust collector (cyclone) outlet, the bag filter inlet, the bag filter outlet, the scrubber inlet, the scrubber outlet, the selective catalytic reduction device inlet, the selective catalytic reduction device outlet, the stack discharge, the water supply tank, the process water storage tank, the caustic soda supply tank, the slaked lime slurry storage tank, the urea solution storage tank, the ammonia storage tank, the diesel storage tank, the scrubber storage tank, the wastewater storage tank, the process water storage tank, the city water storage tank, or the magnesium hydroxide tank, the atmosphere, the warehouse, the hopper section, the primary air supply, the secondary air supply, or the exhaust gas recirculation inlet or The humidity may be related to at least one of the exhaust gas recirculation outlet and the incinerator facility atmospheric humidity. The humidity information may be measured by a humidity sensor included in the sensor unit (320).

[0132] Information about exhaust gas may include information about the amount of at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) emitted from the incinerator. Here, information about the amount may include at least one of molar concentration, mass, or mass concentration. In addition, information about exhaust gas may include information about at least one of flow rate, temperature, pressure, and humidity. Information about exhaust gas may be measured by a material detection sensor or a flow rate, temperature, pressure sensor included in the sensor unit (320). For example, information about exhaust gas may be information acquired from a TMS (Tele-Monitoring System). TMS may refer to a system that constantly measures air pollutants emitted from a chimney of a business establishment with a sensor unit (320) and connects them online to a control center to manage the emission situation in real time. The material detection sensor may be a sensor for detecting a specific chemical substance, and the flow rate, temperature, pressure, and sensors may be sensors for measuring the amount, temperature, pressure, and humidity of a material passing through a specific space. In addition, information on the amount of at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) may represent the mass of the material, but is not limited thereto, and may also represent the molar concentration or mass concentration of the material with respect to the entire material. In addition, information on the exhaust gas may be information on the molar concentration, mass concentration, temperature, pressure, humidity, and flow rate of the material discharged from the combustion chamber (431), the boiler (435), the selective non-catalytic reduction device (421), the centrifugal dust collector (cyclone) (422), the filter dust collector (bag filter) (425), the selective catalytic reduction device (426), the semi-dry scrubber (semi-dry reactor) (423), and the stack (428).

[0133] The operation control device (310) may perform a step (520) of acquiring operation setting information for the analysis time from memory. The operation setting information may refer to parameters directly changed by the user. The user may change the operation status information by modifying the operation setting information. The operation status information may change according to the operation setting information, but may not always change as desired by the user due to various environmental conditions of the incinerator. However, the operation setting information may change as desired by the user unless there is a mechanical / electronic problem.

[0134] More specifically, the operation setting information may include at least one of blower air volume information, pusher information, blower damper opening rate information, chemical input amount information for reducing air pollutants, chemical input amount information for improving incineration efficiency, or waste input setting information. In addition, the operation setting information may include at least one of waste input to the incinerator, blower air volume, blower damper opening rate, chemical input amount for reducing air pollutants, chemical input amount for improving incineration efficiency, pusher (grate) operation cycle or operation number of times. In addition, the blower amount, which is one of the operation setting information, may be controlled by adjusting the output frequency of an electric motor inverter connected to the blower.

[0135] The blower's airflow information may indicate the airflow of a pressurized blower, an induced blower, or an exhaust gas recirculation blower. The pressurized blower may include at least one of a primary pressurized blower or a secondary pressurized blower. Here, the airflow may indicate the output frequency of an electric motor inverter connected to the blower. That is, the operation control device (310) may control the airflow by controlling the output frequency of the electric motor inverter connected to the blower. The airflow information may correspond to the frequency information of the blower.

[0136] The opening degree of a blower damper may refer to the degree of opening of a device that adjusts the amount of air supplied by a pressurized blower, an induced blower, or an exhaust gas recirculation blower. The opening degree of a blower damper may refer to the degree of opening of a damper installed in any section of an air path supplied from a pressurized blower, an induced blower, or an exhaust gas recirculation blower.

[0137] Information about the pusher (grid) may include information about the operating cycle or number of operations of the pusher. Information about the pusher may indicate the operating cycle (speed) or number of operations of the pusher included in the ramp pusher, the drying stage stocker, the combustion stage stocker, or the post-combustion stage stocker. Here, the operation of the pusher may refer to the operation of the pusher pushing the waste. Information about the pusher may also include whether the pusher is running / stopped (on / off). The ramp pusher may be configured to push and move the introduced waste. The waste introduced by the ramp pusher is moved to the drying stage, and may be moved from the drying stage pusher to the combustion stage stocker and then to the post-combustion stage stocker.

[0138] Information on the amount of chemicals to be injected to reduce air pollutants may refer to the amount of urea water, ammonia water, ammonia gas, dilution water, slaked lime, and magnesium hydroxide injected into the polluted gas treatment facility (420) of the incinerator (410) to lower the measured value of at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) measured in the polluted gas treatment facility (420).

[0139] Information on the amount of chemicals injected to improve incineration efficiency may refer to the amount of chemicals injected to optimally change at least one of the operating status information measured in the polluted gas treatment facility (420) of the incinerator (410), the supply flow information, the blower flow information, the pressure information, the information on the storage tank, the temperature information, the humidity information, and the information on the exhaust gas.

[0140] The waste input information may include at least one of the type of waste input into the incinerator, the waste input waiting time (waste input point), the weight (or mass) of the input waste, the predetermined calorific value of the input waste, the predetermined calorific curve of the input waste, the number of times the waste is inputted per day, or the accumulated daily input amount of the waste.

[0141] More specifically, the waste input setting information may include at least one of the waste input point, the input waste weight, and the predetermined heating curve of the input waste. Here, the waste input point can be inversely calculated based on data on the hopper gate waiting time (the waiting time after the first hopper is opened). The hopper gate is a gate for feeding waste into the incinerator. When it is opened, the waste accumulated in the hopper gate is fed into the incinerator, and when it is closed, the waste is not fed into the incinerator. In addition, there may be at least one hopper gate. For example, the incinerator control system (300) may accumulate and store the waiting time after the hopper gate is opened for inverse calculation. The incinerator control system (300) may obtain the next opening point by adding the predetermined waiting time to the opening point.

[0142] Additionally, the incinerator control system (300) may determine the opening point of the primary hopper gate as the waste input point. However, this is not limited thereto. According to various embodiments of the present disclosure, the incinerator control system (300) may obtain the actual input point of waste by adding the following input difference time to the opening point of the primary hopper gate.

[0143] Input differential time = the time from when the first hopper gate opens until the second hopper gate opens and closes and the ramp pusher returns after pushing.

[0144] The feed differential time may vary from incinerator to incinerator. The feed differential time can be changed by administrator input.

[0145] In addition, the weight of the input waste can be inversely calculated using the differential value of the daily accumulated waste input amount. More specifically, the incinerator control system (300) can store the daily accumulated waste input amount over time. The daily accumulated waste input amount can be acquired by the sensor unit (320). The incinerator control system (300) can acquire the weight of the input waste by differentiating the daily accumulated waste input amount. In addition, the incinerator control system (300) can link the inversely calculated weight of the input waste with the input time obtained above. For example, the data of the daily accumulated waste input amount is measured and stored when the waste crane is positioned at the top of the hopper gate and then lowered, and the time when the waste is sprinkled on the hopper gate can be determined based on the data of the rapid decrease in the weight information of the waste crane. The data of the waste input time point closest to the time of sprinklering and the inversely calculated waste weight data can be combined into a single data group. In other words, the incinerator control system (300) can determine the time of the waste input and the weight of the input waste at the time of the waste input.

[0146] The predetermined heating curve of the input waste included in the waste input configuration information may include at least one of the heating point, temperature change curve, and total heating value of the waste. The heating curve may represent the heating value over time when the waste is incinerated. The incinerator control system (300) may obtain the predetermined heating curve of the input waste by performing a reverse calculation using density estimation based on pre-stored statistics. However, the present invention is not limited thereto, and the incinerator control system (300) may obtain the predetermined heating curve using a machine learning model.

[0147] The predetermined calorific value of the input waste can be obtained based on a table that matches the properties of the waste with its calorific value. Alternatively, the calorific value can be automatically obtained based on an image of the input waste using a machine learning model. The predetermined calorific value curve of the input waste can be a curve that shows the calorific value over time from the time of input of a unit weight (or size) of waste. The calorific value curve can also be determined based on a predetermined table or machine learning model.

[0148] The operating status information or operating setting information may further include boiler drum conductivity, information about the boiler drum level control valve, information about the fallout ash transport conveyor, information about the re-discharge conveyor, instantaneous load cell current weight value, and wastewater spray flow rate.

[0149] Information about the boiler drum level control valve may include information about whether the valve is open or closed, or the degree of opening. Furthermore, information about the fallout conveyor and the re-discharge conveyor may include information about whether the conveyor is operating or its speed.

[0150] The incinerator control system (300) can perform a step (530) of acquiring differential information between the current time and the previous time based on operating status information and operating setting information. The step (530) of acquiring differential information is described with reference to FIGS. 6 and 7.

[0151] FIG. 6 is a diagram illustrating an incinerator control system according to one embodiment of the present disclosure. FIG. 7 is a flowchart illustrating the operation of the incinerator control system according to one embodiment of the present disclosure.

[0152] The incinerator control system (300) may perform the following operations to obtain differential information. Referring to FIG. 7, the incinerator control system (300) may perform a step (710) of obtaining at least one of a plurality of pieces of learned operation status information and a plurality of pieces of operation setting information obtained through a basic sampling period. The basic sampling period may be a period in which sensors included in the sensor unit (320) transmit sensor signals to the operation control device (310) or the server (100). The basic sampling period may be changed based on the sensor control signal of the operation control device (310). The basic sampling period may be different for each sensor included in the sensor unit (320). The operation control device (310) may obtain a plurality of pieces of learned operation status information through the basic sampling period. In addition, the operation control device (310) may transmit at least one of the plurality of pieces of learned operation status information and the plurality of pieces of operation setting information obtained through the basic sampling period to the server (100).

[0153] The incinerator control system (300) can obtain operation setting information from memory at a basic sampling cycle. However, this is not limited to this, and the incinerator control system (300) may not obtain operation setting information at a basic sampling cycle. The incinerator control system (300) can obtain operation setting information from memory when necessary.

[0154] The incinerator control system (300) may perform a step (720) of acquiring the current operating status information, the current operating status information, the previous operating status information, and the previous operating status information based on a plurality of pieces of operating status information and a plurality of pieces of operating setting information according to a system cycle. The incinerator control system (300) may resample at least one of the plurality of pieces of operating status information and the plurality of pieces of operating setting information acquired at a basic sampling cycle for each system cycle. The basic sampling cycle may be a value smaller than the system cycle. The incinerator control system (300) may increase the accuracy of the information by resampling at least one of the acquired plurality of pieces of operating status information and the plurality of pieces of operating setting information. This is because although some of the acquired plurality of pieces of operating status information and the plurality of pieces of operating setting information may be erroneous values, the contribution rate of the erroneous values ​​may be greatly reduced by resampling. Accordingly, the accuracy of the incinerator control system (300) may be increased.

[0155] Resampling will be described in detail with reference to FIG. 6. Referring to FIG. 6, the basic sampling period (610) may be shorter than the system period (620). The incinerator control system (300) may acquire at least one of a plurality of pieces of operating status information and a plurality of pieces of operating setting information with the basic sampling period (610). For example, the incinerator control system (300) may acquire a plurality of pieces of information (621, 622, 623, 624, 631, 632, 633, 634). The operation control device may perform resampling with the system period (620).

[0156] The incinerator control system (300) can use the average, median, median, minimum, or maximum value of multiple pieces of information acquired within one system cycle (620) as a resampled value. For example, the incinerator control system (300) can perform a step of acquiring the current time's operating status information, the current time's operating setting information, the previous time's operating status information, and the previous time's operating setting information by averaging multiple pieces of operating status information and multiple pieces of operating setting information for each system cycle. The operating control device (310) can determine the average of the multiple pieces of information (621, 622, 623, 624) included in the previous cycle as the information of the previous time. In addition, the incinerator control system (300) can determine the average of the multiple pieces of information (631, 632, 633, 634) included in the current cycle as the information of the current time. The information of the previous time in FIG. 6 can include the operating status information of the previous time or the operating setting information of the previous time. The current time information in Fig. 6 may include current time driving status information or current time driving setting information.

[0157] The incinerator control system (300) can use the derived values ​​as resampled values ​​by applying multiple pieces of information acquired within a single system cycle (620) to a formula. For example, the operation control device (310) can obtain resampled values ​​by applying Kernel Density Estimation to smooth the PDF (Probability Density Function) of data within the system cycle. In addition, the incinerator control system (300) can use the derived values ​​as resampled values ​​by applying multiple pieces of information acquired within a single system cycle (620) to a machine learning model. According to Kernel Density Estimation, the incinerator control system (300) can store information on the data distribution within the system cycle (620) as continuous information (a discontinuous distribution can be expressed as a continuous distribution). In addition, according to Kernel Density Estimation, the incinerator control system (300) can add the effect of time series (multi-row) prediction (if difference value and gradient (trend) value are macroscopic time series information, Kernel Density Estimation is microscopic time series information) to the cross-sectional value (single row) based prediction. The incinerator control system (300) can set the time period for storing data distribution within the system period and the degree of smoothing for PDF from the user's input. The incinerator control system (300) can perform resampling according to the characteristics of various operation status information or operation setting information included in the incinerator control system (300) by allowing customization of resampling based on the user's input.

[0158] Referring to FIG. 7, the incinerator control system (300) may perform a step (730) of obtaining differential information by subtracting one of the operating status information of the previous time and the operating setting information of the previous time from one of the operating status information of the current time and the operating setting information of the current time. The incinerator control system (300) may obtain differential information by subtracting corresponding pieces of information. For example, the operation control device (310) may obtain differential information by subtracting the operating status information of the previous time from the operating status information of the current time. In addition, the incinerator control system (300) may obtain differential information by subtracting the temperature information of the previous time from the temperature information of the current time. Since the incinerator control system (300) performs resampling at a system cycle (620), the difference between the current time and the previous time may be a system cycle.

[0159] Difference information can include change amount information and change direction information. The size of the difference information corresponds to the change amount information, and the signs of multiple pieces of difference information can correspond to change direction information. Change amount information can refer to the absolute value of the difference obtained by subtracting the information from the previous time from the information from the current time. Furthermore, change direction information can refer to the sign of the difference obtained by subtracting the information from the previous time from the information from the current time.

[0160] Referring again to FIG. 5, the incinerator control system (300) can perform a step (540) of obtaining trend information during the analysis time based on the operating status information and operating setting information.

[0161] Trend information may include differential information. If differential information is the difference between the previous time information and the current time information, the trend information may include the difference between the previous time information and the previous time information. In other words, differential information refers to the most recent information, and trend information may include all differential information within the analysis time period. The incinerator control system (300) may acquire multiple differential information during the analysis time period as trend information.

[0162] Trend information may include information on the amount of change and direction of change of one of the operating status information and operating setting information for each system cycle during the analysis time. The amount of change information may refer to the absolute value of the difference obtained by subtracting the information of the previous time from the information of the current time. In addition, the direction of change information may refer to the sign of the difference obtained by subtracting the information of the previous time from the information of the current time. Trend information may also include information on the amount of change in the difference and information on the direction of change in the difference. In other words, trend information may include information on the amount of change in the slope of the information and information on the direction of change in the slope. Here, the difference may refer to the value obtained by subtracting the information of the previous time from the information of the current time, or the value obtained by subtracting the information of the previous time from the information of the current time.

[0163] The incinerator control system (300) may further perform the following process to perform the step (540) of acquiring trend information. The incinerator control system (300) may perform the step of acquiring a plurality of pieces of operating status information and a plurality of pieces of operating setting information at a basic sampling cycle. The step of acquiring a plurality of pieces of operating status information and a plurality of pieces of operating setting information has already been described in the process of acquiring differential information, so a redundant description will be omitted.

[0164] The incinerator control system (300) may perform a step of acquiring operation status information at an n-th time according to a system cycle, operation setting information at an n-th time, operation status information at an n-1-th time, and operation setting information at an n-1-th time based on a plurality of operation status information and a plurality of operation setting information. The step of acquiring operation status information at an n-th time, operation setting information at an n-th time, operation status information at an n-1-th time, and operation setting information at an n-1-th time may correspond to step (720) of FIG. 7. The difference between the n-th time and the n-1-th time may be a system cycle. Here, n may be a natural number. n may have a number from 1 to N. For example, N may be a value obtained by dividing the analysis time by the system cycle. N may be a natural number.

[0165] The incinerator control system (300) may perform a step of obtaining a plurality of difference information by differentiating one of the operation status information at the n-1th time and the operation setting information at the n-1th time from one of the operation status information at the n-th time and the operation setting information at the n-1th time. This step may correspond to step (730) of FIG. 7. The incinerator control system (300) may perform a step of obtaining trend information of the plurality of difference information. The operation control device (310) may obtain difference information by subtracting corresponding information. For example, the incinerator control system (300) may obtain difference information by subtracting the operation status information at the n-1th time from the operation status information at the n-th time. In addition, the incinerator control system (300) may obtain difference information by subtracting the temperature information at the n-1th time from the temperature information at the n-th time. Since the incinerator control system (300) performs resampling in a system cycle (620), the difference between the nth time and the n-1th time may be a system cycle.

[0166] The size of the plurality of difference information may correspond to the change amount information of the trend information, and the sign of the plurality of difference information may correspond to the change direction information of the trend information.

[0167] Referring back to FIG. 5, the incinerator control system (300) may perform a step (550) of obtaining predicted operation control information by applying at least one of operation status information, operation setting information, difference information, and trend information to the operation simulation machine learning model (820). The operation control device (310) according to one embodiment of the present disclosure may perform a step (550) of obtaining predicted operation control information by applying at least one of operation status information, operation setting information, difference information, and trend information to the operation simulation machine learning model (820) received from the server (100). The operation status information and operation setting information input to the operation simulation machine learning model may be values ​​obtained by resampling the signal of the sensor unit (320) at a system sampling cycle. The operation simulation machine learning model (820) is described in FIG. 8. Step (550) may be performed in the data recognition unit (120) of FIG. 1 included in the server (100) or the operation control device (310).

[0168] Trend information may include differential information. However, the driver simulation machine learning model (820) can utilize both differential information and trend information. This is because differential information is obtained using the most recent information, and thus, differential information must be input separately into the driver simulation machine learning model (820) to apply different weights to it than trend information. However, this is not limited to this, and the driver simulation machine learning model (820) can also utilize only trend information as input.

[0169] In addition, the types of information included in the operating status information and the operating setting information may be different from the types of information included in the differential information and trend information. The incinerator control system (300) may utilize differential information or trend information for some types of information among the multiple types of information included in the operating status information and the operating setting information. The types of information may each indicate the mass information of waste, the supply flow information, the blower flow information, the pressure information, the information about the storage tank, the temperature information, the humidity information, and the information about the exhaust gas included in the operating status information. In addition, the types of information may each indicate the information such as whether waste is input, the blower blower volume, the opening rate of the blower damper, the amount of chemicals injected to reduce air pollutants, the amount of chemicals injected to improve incineration efficiency, and the operation cycle or number of operations of the pusher (grate) in the operating setting information. In addition, the airflow amount, which is one of the operation setting information, can represent the output frequency of the electric motor inverter connected to the blower. That is, the driver simulation machine learning model (820) can receive as input at least one of the following information: mass information of waste, supply flow information, airflow information, pressure information, information about storage tank, temperature information, humidity information, and exhaust gas information included in the operation status information. In addition, the driver simulation machine learning model (820) can receive as input at least one of the following information: whether waste is input, the airflow amount of the blower, the opening rate of the blower damper, the amount of chemicals injected to reduce air pollutants, the amount of chemicals injected to improve incineration efficiency, the operation cycle or number of operations of the pusher (grate). In addition, the airflow amount, which is one of the operation setting information, can receive as input the output frequency of the electric motor inverter connected to the blower. In addition, the driver simulation machine learning model (820) can receive as input the mass information of waste, supply flow information, blower flow information, pressure information, information about storage tanks, temperature information, humidity information, and differential information or trend information about exhaust gas.

[0170] The type of information included in the driving control information may be the same as the type of information included in the driving setting information, or may be part of the types of information included in the driving setting information. The predicted driving control information output from the driver simulation machine learning model (820) may include one of the following: predicted information on whether or not waste is input, predicted weight of input waste, predicted calorific value of input waste, predicted calorific curve of input waste, predicted air volume of an induced blower, predicted air volume of a pressure blower, predicted exhaust gas recirculation blower, predicted opening rate information of a blower damper, predicted chemical input amount for reducing air pollutants, predicted chemical input amount for improving incineration efficiency, or predicted pusher operation information. The air volume may include information on the predicted output frequency of an electric motor inverter connected to the blower.

[0171] The predicted waste input information may indicate at least one of the input amount or input status of waste. The predicted airflow information of the induced blower may include information on the predicted output frequency of the electric motor inverter connected to the induced blower. In addition, the predicted airflow information of the pressurized blower may include information on the predicted output frequency of the electric motor inverter connected to the pressurized blower. In addition, the predicted airflow of the exhaust gas recirculation blower may include information on the predicted output frequency of the electric motor inverter connected to the exhaust gas recirculation blower. The airflow may indicate information on the predicted output frequency of the electric motor inverter connected to the blower.

[0172] In addition, the predicted opening rate information of the blower damper may indicate the predicted opening rate of a damper installed at least one of an induced blower, a pressure blower, and an exhaust gas recirculation blower outlet. In addition, the predicted chemical injection amount information for reducing air pollutants may include at least one of the type, injection timing, and injection amount of the injected chemical. In addition, the predicted chemical injection amount information for improving incineration efficiency may include at least one of the type, injection timing, and injection amount of the injected chemical. In addition, the predicted pusher operation information may include at least one of the cycle (speed) at which the pusher pushes the waste or whether the pusher is operated (on / off).

[0173] The incinerator control system (300) may further perform a step of controlling the incinerator based on predicted operation control information. The operation control device (310) may control the incinerator based on one of the predicted information on whether or not waste is input, the predicted blowing amount of an induced blower, the predicted blowing amount of a pressure blower, the predicted blowing amount of an exhaust gas recirculation blower, the predicted opening rate information of a blower damper, the predicted chemical injection amount for reducing air pollutants, the predicted chemical injection amount for improving incineration efficiency, or the predicted pusher operation information. The blowing amount may be controlled by using the predicted output frequency of an electric motor inverter connected to the blower as information.

[0174] The driving control device (310) may further perform a step of displaying predicted driving control information. For example, the driving control device (310) may perform a step of displaying at least one of predicted information on whether or not waste is input, predicted information on the blowing volume of an induced blower, predicted information on the blowing volume of a pressure blower, predicted information on the blowing volume of an exhaust gas recirculation blower, predicted information on the opening rate of a blower damper, predicted information on the amount of chemicals injected to reduce air pollutants, predicted information on the amount of chemicals injected to improve incineration efficiency, or predicted information on the operation of a pusher. The user may automatically check the controlled value.

[0175] More specifically, the predicted operation control information received from the driver simulation machine learning model (820) may indicate predicted information on whether or not to input waste. The operation control device (310) may further perform the following process to perform a step of controlling the incinerator based on the predicted operation control information. If the predicted information on whether or not to input waste indicates that waste is input, the operation control device (310) may perform a step of inputting waste. In addition, if the predicted information on whether or not to input waste indicates that waste is not input, the operation control device (310) may perform a step of not inputting waste.

[0176] In addition, the operation control device (310) can control at least one of the rotation, rotation speed, and damper opening rate of the induced blower, the pressure blower, and the exhaust gas recirculation blower based on the predicted blowing amount of the induced blower, the predicted blowing amount of the pressure blower, or the predicted blowing amount of the exhaust gas recirculation blower included in the operation control information. In addition, the operation control device (310) can control at least one of the predicted chemical injection amount for reducing air pollutants and the predicted chemical injection amount for improving incineration efficiency based on the exhaust gas number information included in the operation control information. In addition, the operation control device (310) can control at least one of the operation and operation speed of the pusher based on the predicted pusher operation information included in the operation control information.

[0177] However, it is not limited thereto. The operation control device (310) may not automatically control the incinerator based on the predicted operation control information. The operation control device (310) may only display at least one of the predicted information on whether or not waste is input, the predicted blowing amount of the induced blower, the predicted blowing amount of the pressurized blower, the predicted blowing amount of the exhaust gas recirculation blower, the predicted opening rate information of the blower damper, the predicted chemical injection amount to reduce air pollutants, the predicted chemical injection amount to improve incineration efficiency, or the predicted pusher operation information. The user may input an input to the operation control device (310) for controlling the incinerator based on the presented predicted operation control information. The operation control device (310) may control the incinerator based on the input received from the user.

[0178] FIG. 8 is a diagram illustrating a driver simulation machine learning model according to one embodiment of the present disclosure.

[0179] First, the process of generating a driver simulation machine learning model (820) will be described. The step of generating a driver simulation machine learning model (820) can be performed in the data learning unit (110) of FIG. 1 included in the server. The server (100) can perform a step of acquiring a plurality of learning driving status information (811), a plurality of learning driving setting information (812), and a plurality of driving control information (813) from a pre-collected learning database. Here, the plurality of learning driving status information (811) and the plurality of learning driving setting information (812) are learning data, and the plurality of driving control information (813) may be label information. The plurality of driving control information (813) may be ground truth information. That is, the plurality of driving control information (813) may be driving control information actually input by a skilled user.

[0180] The incinerator control system (300) can obtain a plurality of pieces of learning operation status information (811), a plurality of pieces of learning operation setting information (812), and a plurality of pieces of operation control information (813) as follows. The incinerator control system (300) can continuously collect user input, operation status information, and operation setting information. In addition, when the user inputs operation control information for controlling the incinerator, the incinerator control system (300) can obtain the learning operation status information and the learning operation setting information for the previous analysis time from the memory. That is, the incinerator control system (300) can obtain the learning operation status information and the learning operation setting information corresponding to the operation control information from the memory. In addition, using this method, the incinerator control system (300) can obtain a plurality of pieces of learning operation status information (811) and a plurality of pieces of learning operation setting information (812) corresponding to the plurality of pieces of operation control information (813).

[0181] The operation control information actually entered by the skilled user may be determined based on the experience of the skilled user, but the incinerator control system (300) may also automatically select from among multiple operation control information of multiple users. For example, the incinerator control system (300) may obtain multiple candidate learning operation status information, multiple candidate learning operation setting information, and multiple candidate operation control information. The multiple candidate learning operation status information, the multiple candidate learning operation setting information, and the multiple candidate operation control information may be data obtained at the incinerator site. The multiple candidate learning operation status information, the multiple candidate learning operation setting information, and the multiple candidate operation control information may be data at a point in time at the incinerator site.

[0182] The incinerator control system (300) can obtain operation status information corresponding to a plurality of candidate learning operation status information, a plurality of candidate learning operation setting information, and a plurality of candidate operation control information after a predetermined time. That is, the incinerator control system (300) can obtain operation status information corresponding to candidate learning operation status information, selected candidate learning operation setting information, and selected candidate operation control information selected from a plurality of candidate learning operation status information, a plurality of candidate learning operation setting information, and a plurality of candidate operation control information after a predetermined time. Here, the operation status information after a predetermined time may refer to operation status information after a predetermined time after the incinerator is operated based on the candidate learning operation status information, the candidate learning operation setting information, and the candidate operation control information.

[0183] Operating status information may include temperature, nitrogen oxides, carbon monoxide, hydrogen chloride, sulfur dioxide, dust, and the amount of waste discharged per hour. The predetermined time may be the same as the predicted time. For example, the predicted time may be greater than or equal to 1 second and less than or equal to 4 hours. Furthermore, the predicted time may be greater than or equal to 1 second and less than or equal to 4 hours. However, this is not a limitation.

[0184] The incinerator control system (300) can obtain skill information (r_t) at time t based on operating status information after a predetermined time.

[0185] r_t = w1 * r_t_temp + w2 * r_t_NOX + w3 * r_t_CO + w4 * r_t_HCL + w5 * r_t_SOX + w6 * r_t_Dust + w7 * r_t_waste

[0186] Here, w1 to w7 are predetermined weights and may be positive real numbers. Here, r_t_temp may be a reward value according to the learning temperature information at time t. The learning temperature information may be temperature information at the outlet of the incinerator. r_t_NOX is a reward value according to the amount of learning nitrogen oxide, r_t_CO is a reward value according to the amount of learning carbon monoxide, r_t_HCL is a reward value according to the amount of learning hydrogen chloride, r_t_SOX is a reward value according to the amount of learning sulfur oxide, and r_t_Dust may be a reward value according to the amount of learning dust. In addition, r_t_waste may be a reward value according to the amount (weight) of waste input per learning time. The incinerator control system (300) may store a table storing a predetermined reward value according to the learning temperature information. The incinerator control system (300) may store a table storing a predetermined reward value according to the amount of learning nitrogen oxide. The incinerator control system (300) may store a table storing predetermined compensation values ​​according to the amount of learned carbon monoxide. The incinerator control system (300) may store a table storing predetermined compensation values ​​according to the amount of learned hydrogen chloride. The incinerator control system (300) may store a table storing predetermined compensation values ​​according to the amount of learned sulfur oxide. The incinerator control system (300) may store a table storing predetermined compensation values ​​according to the amount of learned dust. The incinerator control system (300) may store a table storing predetermined compensation values ​​according to the amount (weight) of waste input per learning time.

[0187] The proficiency information (r_t) at time t is not limited to the above formula, and at least one of r_t_temp, r_t_NOX, r_t_CO, r_t_HCL, r_t_SOX, r_t_Dust, and r_t_waste may be omitted. The incinerator control system (300) can obtain proficiency information (r_t) for a time range. For example, the time range may be a system sampling period. The incinerator control system (300) can obtain proficiency information (R) for a time range (from time t to time t+alpha) based on the proficiency information at time t.

[0188] R = r_t + r_(t+1) + ... + r_(t+alpha)

[0189] Here, the difference between time t and time t+1 may be equal to the basic sampling period. However, it is not limited thereto. If the proficiency information (R) for the time range is greater than or equal to a predetermined threshold compensation value, the incinerator control system (300) may determine the candidate operation control information for the time range as that of the skilled user. That is, if the proficiency information (R) for the time range is greater than or equal to a predetermined threshold compensation value, the incinerator control system (300) may determine the candidate operation control information for the time range as one of the plurality of operation control information (813). In addition, if the proficiency information (R) for the time range is greater than or equal to a predetermined threshold compensation value, the incinerator control system (300) may collect from the memory a plurality of candidate learning operation status information and a plurality of candidate learning operation setting information during the previous analysis time of the candidate operation control information. The incinerator control system (300) may perform resampling on the plurality of candidate learning operation status information and the plurality of candidate learning operation setting information for the time range. In addition, the incinerator control system (300) can determine the resampled candidate learning operation status information as one of the plurality of learning operation status information (811) and determine the resampled candidate learning operation setting information as one of the plurality of learning operation setting information (812). The incinerator control system (300) can repeat the above process to obtain a plurality of learning operation status information (811), a plurality of learning operation setting information (812), and a plurality of operation control information (813). The incinerator control system (300) can generate an operator simulation machine learning model using the plurality of learning operation status information (811), the plurality of learning operation setting information (812), and the plurality of operation control information (813). Since resampling has already been described, a redundant description will be omitted.

[0190] The incinerator control system (300) may determine that the operation control information for the time range belongs to an unskilled user if the proficiency information (R) for the time range is less than a predetermined threshold compensation value. The incinerator control system (300) may not perform resampling for the time range and may not use the time range for generating a machine learning model if the proficiency information (R) for the time range is less than a predetermined threshold compensation value.

[0191] The incinerator control system (300) can obtain a plurality of resampled learning operation status information and a plurality of learning operation setting information based on the above process.

[0192] The incinerator control system (300) can generate a driver simulation machine learning model (820) based on a plurality of resampled learning operation status information and a plurality of learning operation setting information. The incinerator control system (300) can perform a step of acquiring a plurality of past difference information and a plurality of past trend information based on a plurality of learning operation status information and a plurality of learning operation setting information. The process of acquiring a plurality of past difference information and a plurality of past trend information has been described in FIG. 7, so a redundant description will be omitted.

[0193] The incinerator control system (300) can perform a step of generating a driver simulation machine learning model (820) by machine learning the correlation (causality) of a plurality of driving control information for at least one of a plurality of learning driving status information, a plurality of learning driving setting information, a plurality of past difference information, and a plurality of past trend information. The incinerator control system (300) can increase the accuracy of the driver simulation machine learning model (820) by updating the weights included in the driver simulation machine learning model (820) by performing forward propagation and back propagation.

[0194] For example, a driver simulation machine learning model (820) can be generated in a server (100). The server (100) can store the driver simulation machine learning model (820) in memory. In addition, the server (100) can perform a step of transmitting the driver simulation machine learning model (820) to a driving control device (310) or another server. The server (100) or the driving control device (310) can obtain predicted driving control information (840) using the driver simulation machine learning model (820).

[0195] As described in FIG. 5, the driving control device (310) can newly collect driving status information (831) and driving setting information (832) as in steps (510) and (520). Here, the driving status information (831) and driving setting information (832) collected by the driving control device (310) may be future information compared to the plurality of learning driving status information (811), the plurality of learning driving setting information (812), and the plurality of driving control information (813) used when generating the driver simulation machine learning model (820). The driving status information (831) and the driving setting information (832) may be resampled values ​​obtained from the sensor unit (320). It should be noted that the driving control device (310) or the server (100) may use the driving status information (831) and the driving setting information (832) to update the driver simulation machine learning model (820).

[0196] The incinerator control system (300) can obtain differential information and trend information, such as steps (530) and (540). In addition, the incinerator control system (300) can perform step (550) of obtaining predicted operation control information (840) by applying at least one of operation status information, operation setting information, differential information, and trend information to a driver simulation machine learning model (820). Step (550) can be performed in the data recognition unit (120) of FIG. 1 included in the server (100) or the operation control device (310).

[0197] The incinerator control system (300) of the present disclosure generates predicted operation control information (840) based on a driver simulation machine learning model (820), thereby assisting a user in determining operation control information based on the predicted operation control information (840). Furthermore, even if the incinerator user is not an expert, the system can help maintain the incinerator in optimal condition.

[0198] FIG. 9 is a flowchart illustrating the operation of an incinerator control system according to another embodiment of the present disclosure.

[0199] The incinerator control system (300) may include a sensor unit (320), an operation control device (310), and a server (100). The incinerator control system may control the incinerator using machine learning. The operation of the incinerator control system is described below. Steps (510) to (540) of FIG. 9 may correspond to steps (510) to (540) of FIG. 5. In the description of FIG. 9, redundant descriptions of matters already described in FIG. 5 will be omitted.

[0200] The driving control device (310) can perform a step (510) of acquiring driving status information for a predetermined analysis time from the sensor unit (320). The driving control device (310) can perform a step (520) of acquiring driving setting information for the analysis time from the memory. The incinerator control system (300) can perform a step (530) of acquiring difference information between the current time and the previous time based on the driving status information and the driving setting information. In addition, the incinerator control system (300) can perform a step (540) of acquiring trend information for the analysis time based on the driving status information and the driving setting information.

[0201] The incinerator control system (300) may perform a step (950) of obtaining predicted operating status information by applying at least one of operating status information, operating setting information, difference information, and trend information to the operating status prediction machine learning model (1020). For example, the operating control device (310) may perform a step (950) of obtaining predicted operating status information by applying at least one of operating status information, operating setting information, difference information, and trend information to the operating status prediction machine learning model (1020) received from the server. The operating status information and operating setting information input to the operating status prediction machine learning model (1020) may be values ​​obtained by resampling the signal of the sensor unit (320) at a system sampling cycle. The operating status prediction machine learning model (1020) is described in FIG. 10. Step (950) may be performed in the data recognition unit (120) of FIG. 1 included in the server (100) or the operating control device (310).

[0202] Trend information may include differential information. However, the driving condition prediction machine learning model (1020) can utilize both differential information and trend information. This is because differential information is obtained using the most recent information, and thus, differential information must be input separately into the driving condition prediction machine learning model (1020) to apply different weights to the differential information than to the trend information. However, this is not limited to this, and the driving condition prediction machine learning model (1020) can also utilize only trend information as input.

[0203] In addition, the types of information included in the driving status information and driving setting information input to the driving status prediction machine learning model (1020) and the types of information included in the differential information and trend information may be different. The driving control device (310) may use differential information or trend information for some types of information among the multiple types of information included in the driving status information and driving setting information. The types of information may each represent waste mass information, supply flow information, blower flow information, pressure information, information about storage tanks, temperature information, humidity information, and exhaust gas information included in the driving status information. In addition, the types of information may each represent the blower volume of the blower damper, the opening rate of the blower damper, the amount of chemicals injected to reduce air pollutants, the amount of chemicals injected to improve incineration efficiency, information about the pusher (grate), or waste input information of the incinerator included in the driving setting information. The blower volume may represent the output frequency information of the electric motor inverter connected to the blower. That is, the driving state prediction machine learning model (1020) can receive as input at least one of the following: mass information of waste, supply flow rate information, blower flow rate information, pressure information, information about storage tank, temperature information, humidity information, and exhaust gas information included in the driving state information. In addition, the driving state prediction machine learning model (1020) can receive as input at least one of the following: blower air volume, blower damper opening rate, chemical input amount for reducing air pollutants, chemical input amount for improving incineration efficiency, pusher (grate) information, or waste input information of the incinerator. In addition, the driving state prediction machine learning model (1020) can receive as input the following: mass information of waste, supply flow rate information, blower flow rate information, pressure information, information about storage tank, temperature information, humidity information, and differential information or trend information about exhaust gas.

[0204] The predicted driving status information may be driving status information at a future time after a predetermined prediction time from the current time. That is, the predicted driving status information may be predicted information after a predetermined prediction time from the driving status information input to the driving status prediction machine learning model (1020). The predicted time may be a value of 1 second or more and 4 hours or less. The predicted driving status information is a value predicted by the driving status prediction machine learning model (1020). In addition, the predicted driving status information may indicate the predicted driving status information when the driving setting information input to the driving status prediction machine learning model (1020) is maintained for the predicted time. The predicted driving status information may include at least one of predicted waste mass information, predicted supply flow information, predicted blower flow information, predicted pressure information, predicted storage tank information, predicted temperature information, predicted humidity information, and predicted exhaust gas information.

[0205] The driving control device (310) can perform a step of displaying predicted driving status information. The user can check the predicted driving status information. Based on the predicted driving status information, the user can determine the action to be taken. According to the incinerator control system of the present disclosure, it can help the user quickly predict future situations and operate the incinerator within a normal range.

[0206] The driving control device (310) can also generate predicted driving control information based on current driving status information and driving status information after a predicted time. Accordingly, the driving control device (310) can automatically control the incinerator based on the predicted driving control information.

[0207] When the operation control device (310) performs the step of outputting the predicted operation status information, the operation control device (310) may further perform the following process. The operation control device (310) may perform a step of determining whether the predicted operation status information is within the range of predetermined critical environment information. Here, the predicted operation status information is information obtained from the operation status prediction machine learning model (1020) and may include at least one of predicted waste mass information, predicted supply flow information, predicted blower flow information, predicted pressure information, predicted storage tank information, predicted temperature information, predicted humidity information, and predicted exhaust gas. In addition, the critical environment information may include at least one of critical waste mass information, critical supply flow information, critical blower flow information, critical pressure information, information about the inlet storage tank, critical temperature information, critical humidity information, and critical exhaust gas. In addition, the critical environment information may include at least one of critical temperature range information, critical oxygen amount range information, and critical pressure range information. Critical environmental information may be a range of parameters that an incinerator must satisfy. Critical environmental information may be predetermined by legal standards or the incinerator company's operating guidelines.

[0208] The operation control device (310) can automatically perform control when the predicted operation status information is not included in the critical environment information. The operation control device (310) can automatically generate predicted operation control information when the predicted operation status information is not included in the critical environment information. The operation control device (310) can pre-store an operation control information generation table that corresponds the types of unsatisfied critical environment information to the predicted operation control information. When the predicted operation status information is not included in the critical environment information, the operation control device (310) can automatically generate the predicted operation control information based on the table. In addition, the incinerator control system can automatically control the incinerator to satisfy the critical environment information.

[0209] For example, the operation control device (310) may automatically perform control when the predicted temperature information included in the predicted operation status information is not included in the critical temperature range information included in the critical environment information. More specifically, when the predicted temperature information is lower than or equal to the lower limit of the critical temperature range information, the operation control device (310) may perform a step of generating information indicating that waste should be input based on the operation control information generation table. The operation control information generation table may include predicted operation control information when the predicted operation status information is lower than the lower limit of the critical temperature range information or predicted operation control information when the predicted temperature range information exceeds the upper limit of the critical temperature range information. The operation control device (310) may control the incinerator based on the predicted operation control information, thereby satisfying the critical environment information.

[0210] The operation control device (310) can output information indicating whether to input the generated waste. The user can decide whether to input the waste based on the information indicating whether to input the waste displayed on the operation control device (310) and input the information into the operation control device (310). Alternatively, the operation control device (310) can automatically input the waste based on the information indicating whether to input the generated waste.

[0211] If the predicted temperature information included in the predicted driving state information is included in the critical temperature range information included in the critical environment information, the driving control device (310) can output that the predicted driving state information is included in the critical environment information. The user can confirm that the predicted driving state information satisfies the critical environment information. The above description is based on the critical temperature range information included in the critical environment information, but the same description is possible for the critical oxygen amount range and critical pressure range information, and redundant descriptions are omitted.

[0212] If the operating status information of an incinerator deviates from the range of critical environmental information, the incinerator's incineration capacity may be impaired, or incomplete combustion may result in the generation of hazardous substances. The incinerator control system (300) of the present disclosure can control the operating status information to remain within the range of critical environmental information, thereby maintaining the incinerator in an optimal state for incineration. Furthermore, it can assist in maintaining the incinerator in an optimal state even for non-expert users.

[0213] FIG. 10 is a diagram for explaining a driving state prediction machine learning model according to one embodiment of the present disclosure.

[0214] First, the process of generating a driving state prediction machine learning model (1020) will be described. The step of generating a driving state prediction machine learning model (1020) can be performed in the data learning unit (110) of FIG. 1 included in the server. The server (100) can perform a step of acquiring a plurality of pieces of learning driving state information (1011), a plurality of pieces of learning driving setting information (1012), and a plurality of pieces of driving state information (1013) after a prediction time from a previously collected learning database. Here, the plurality of pieces of learning driving state information (1011) and the plurality of pieces of learning driving setting information (1012) are learning data, and the driving state information (1013) after a plurality of prediction times may be label information. The driving state information (1013) after a plurality of prediction times may be ground truth information. That is, the driving state information (1013) after a plurality of prediction times may be a value acquired from the sensor unit (320).

[0215] The incinerator control system (300) can obtain a plurality of pieces of learning operation status information (1011), a plurality of pieces of learning operation setting information (1012), and a plurality of pieces of operation status information (1013) after a prediction time as follows. The incinerator control system (300) can continuously collect user input, operation status information, and operation setting information. In addition, the incinerator control system (300) can collect a plurality of pieces of learning operation status information (1011) and a plurality of pieces of learning operation setting information (1012) from a memory for a predetermined analysis time. In addition, the operation status information (1013) after a plurality of prediction times can be information after a prediction time after obtaining the plurality of pieces of learning operation status information (1011). The server (100) or the operation control device (310) can obtain the operation status information (1013) after a plurality of prediction times after a predetermined prediction time from data at the last time of the plurality of pieces of learning operation status information (1011).

[0216] Each of the plurality of pieces of learning driving status information may be resampled information. Furthermore, each of the plurality of pieces of learning driving configuration information may be resampled information. Since resampling has already been described, a further explanation will be omitted. Each of the driving status information (1013) after a plurality of prediction times may be resampled information.

[0217] The incinerator control system (300) can generate an operation state prediction machine learning model (1020) based on a plurality of resampled learning operation state information and a plurality of learning operation setting information. The incinerator control system (300) can perform a step of acquiring a plurality of past difference information and a plurality of past trend information based on a plurality of learning operation state information and a plurality of learning operation setting information. The process of acquiring a plurality of past difference information and a plurality of past trend information has been described in FIG. 7, so a redundant description will be omitted.

[0218] The incinerator control system (300) can perform a step of generating an operation state prediction machine learning model (1020) by machine learning the correlation (causality) of operation state information (1013) after a plurality of prediction times for at least one of a plurality of learned operation state information (1011), a plurality of learned operation setting information (1012), a plurality of past difference information, and a plurality of past trend information. The incinerator control system (300) can increase the accuracy of the operation state prediction machine learning model (1020) by updating the weights included in the operation state prediction machine learning model (1020) by performing forward propagation and back propagation.

[0219] For example, a driving state prediction machine learning model (1020) can be generated in a server (100). The server (100) can store the driving state prediction machine learning model (1020) in memory. In addition, the server (100) can perform a step of transmitting the driving state prediction machine learning model (1020) to a driving control device (310) or another server. The server (100) or the driving control device (310) can obtain predicted driving state information (1040) using the driving state prediction machine learning model (1020).

[0220] As described in FIG. 9, the driving control device (310) can newly collect driving status information (1031) and driving setting information (1032) as in steps (510) and (520). Here, the driving status information (1031) and driving setting information (1032) collected by the driving control device (310) may be future information compared to the plurality of learning driving status information (1011), the plurality of learning driving setting information (1012), and the driving status information (1013) after the plurality of predicted times used when generating the driving status prediction machine learning model (1020). The driving status information (1031) and the driving setting information (1032) may be obtained by resampling the plurality of pieces of information obtained from the sensor unit (320). It should be noted that the incinerator control system (300) can additionally use operating status information (1031) and operating setting information (1032) to further update the operating status prediction machine learning model (1020).

[0221] The incinerator control system (300) can obtain differential information and trend information, such as steps (530) and (540) of FIG. 9. Furthermore, the incinerator control system (300) can perform step (950) of applying at least one of the operating status information, the operating setting information, the differential information, and the trend information to the operating status prediction machine learning model (1020) to obtain predicted operating status information (1040). Step (950) can be performed in the data recognition unit (120) of FIG. 1 included in the server (100) or the operating control device (310).

[0222] The incinerator control system (300) of the present disclosure generates operating status information (1040) predicted by an operating status prediction machine learning model (1020), thereby assisting a user in checking the operating status information (1040) and determining operating control information. Furthermore, even if the incinerator user is not an expert, the system can help maintain the incinerator in optimal condition.

[0223] FIG. 11 is a flowchart illustrating an operation method of an incinerator control system according to another embodiment of the present disclosure.

[0224] The incinerator control system (300) may include a sensor unit (320), an operation control device (310), and a server (100). The incinerator control system (300) may control the incinerator using a reinforcement learning model. The operation of the incinerator control system is described below. Steps (510) to (540) of FIG. 11 may correspond to steps (510) to (540) of FIG. 5. Therefore, redundant descriptions of matters already described in FIG. 5 will be omitted.

[0225] The driving control device (310) can perform a step (510) of acquiring driving status information for a predetermined analysis time from the sensor unit. The driving control device (310) can perform a step (520) of acquiring driving setting information for the analysis time from the memory. The incinerator control system (300) can perform a step (530) of acquiring difference information between the current time and the previous time based on the driving status information and the driving setting information. The incinerator control system (300) can perform a step (540) of acquiring trend information for the analysis time based on the driving status information and the driving setting information.

[0226] The incinerator control system (300) can perform a step (1110) of determining at least one of operation status information, operation setting information, difference information, and trend information as current status information. The agent of the action of the incinerator control reinforcement learning model can be the operation control device (310) or the server (100). In addition, the incinerator control reinforcement learning model can determine the current status information using at least one of the operation status information, operation setting information, difference information, and trend information. The incinerator control system (300) can perform an optimal action to obtain an optimal reward from the current status information. The optimal action may vary depending on the current status information. Therefore, the incinerator control system (300) can check the current status information. The incinerator control system (300) can apply the current status information to the incinerator control reinforcement learning model to determine an action to obtain the optimal reward.

[0227] The incinerator control system (300) can perform a step (1120) of applying current state information to an incinerator control reinforcement learning model to obtain predicted operation control information that provides maximum compensation. The operation control device (310) can perform a step (1120) of applying current state information to an incinerator control reinforcement learning model received from a server (100) to obtain predicted operation control information that provides maximum compensation. Step (1120) can be performed in the data recognition unit (120) of FIG. 1 included in the server (100) or the operation control device (310).

[0228] Trend information contained in current state information may include differential information. However, the incinerator control reinforcement learning model can utilize both differential and trend information. This is because differential information is obtained using the most recent information, so it must be input separately into the incinerator control reinforcement learning model to apply different weights to it than trend information. However, this is not limited to this approach; the incinerator control reinforcement learning model can also utilize only trend information as input.

[0229] In addition, the types of information included in the operation status information and operation setting information input to the incinerator control reinforcement learning model may be different from the types of information included in the differential information and trend information. The incinerator control system (300) may utilize differential information or trend information for some types of information among the multiple types of information included in the operation status information and operation setting information. The types of information may each represent waste mass information, supply flow information, blower flow information, pressure information, information about storage tanks, temperature information, humidity information, and exhaust gas information included in the operation status information. In addition, the types of information may each represent the blower air volume, the opening rate of the blower damper, the amount of chemicals injected to reduce air pollutants, the amount of chemicals injected to improve incineration efficiency, the operation cycle or number of operations of the pusher (grate), or information on whether waste is injected into the incinerator included in the operation setting information. That is, the incinerator control reinforcement learning model can receive as input at least one of the following: mass information of waste, supply flow rate information, blower flow rate information, pressure information, information about storage tank, temperature information, humidity information, and exhaust gas information included in the operation status information. In addition, the incinerator control reinforcement learning model can receive as input at least one of the following: blower volume, blower damper opening rate, chemical input amount for reducing air pollutants, chemical input amount for improving incineration efficiency, pusher (grate) operation cycle or number of operations, or information on whether waste is input to the incinerator. In addition, the incinerator control reinforcement learning model can receive as input the following: mass information of waste, supply flow rate information, blower flow rate information, pressure information, information about storage tank, temperature information, humidity information, and differential information or trend information about exhaust gas.

[0230] The type of information included in the operation control information may be the same as the type of information included in the operation setting information, or may be part of the types of information included in the operation setting information. It may include one of the predicted waste mass information, predicted supply flow information, predicted blower flow information, predicted pressure information, predicted storage tank information, predicted temperature information, predicted humidity information, and predicted exhaust gas information output from the incinerator control reinforcement learning model. The predicted waste input information may indicate the input amount and whether or not the waste is input. The predicted blower volume information of the induced blower may indicate the blower volume of the induced blower. In addition, the predicted blower volume information of the pressurized blower may indicate the blower volume of the pressurized blower. In addition, the predicted blower volume of the exhaust gas recirculation blower may indicate the blower volume of the exhaust gas recirculation blower. The blower volume may indicate the predicted output frequency of the electric motor inverter connected to the blower. In addition, the predicted blower damper opening rate information may indicate the opening rate of the damper installed at the outlet of the induced blower, the pressurized blower, and the exhaust gas recirculation blower. In addition, the predicted chemical injection amount information for reducing air pollutants may indicate at least one of the type, timing, and amount of the injected chemical. In addition, the predicted chemical injection amount information for improving incineration efficiency may indicate at least one of the type, timing, and amount of the injected chemical. In addition, the predicted pusher operation information may indicate at least one of the cycle (speed) at which the pusher pushes the waste or whether the pusher is operated (on / off). The incinerator control system (300) may further perform a step of controlling the incinerator based on the predicted operation control information.That is, the operation control device (310) may include one of the following: predicted waste input information, predicted blowing volume of the induced blower, predicted blowing volume of the pressurized blower, predicted exhaust gas recirculation blower blowing volume, predicted blower damper opening rate information, predicted chemical injection amount for reducing air pollutants, predicted chemical injection amount for improving incineration efficiency, or predicted pusher operation information. The blowing volume may include information on the predicted output frequency of the electric motor inverter connected to the blower. The operation control device (310) may further perform a step of displaying the predicted operation control information. For example, the driving control device (310) may perform a step of displaying at least one of the following: information on whether or not waste has been input, the predicted blowing volume of the induced blower, the predicted blowing volume of the pressurized blower, the predicted blowing volume of the exhaust gas recirculation blower, the predicted opening rate information of the blower damper, the predicted chemical injection amount for reducing air pollutants, the predicted chemical injection amount for improving incineration efficiency, or the predicted pusher operation information. The blowing volume may perform a step of displaying information on the predicted output frequency of the electric motor inverter connected to the blower. The user may check the automatically controlled value.

[0231] More specifically, the predicted operation control information received from the incinerator control reinforcement learning model may indicate predicted information on whether or not waste is input. The incinerator control system (300) may further perform the following process to perform a step of controlling the incinerator by the operation control device based on the predicted operation control information. If the predicted information on whether or not waste is input indicates that waste is input, the operation control device (310) may perform a step of inputting waste. In addition, if the predicted information on whether or not waste is input indicates that waste is not input, the operation control device (310) may perform a step of not inputting waste.

[0232] However, it is not limited thereto. The operation control device (310) may not automatically control the incinerator based on the predicted operation control information. The operation control device (310) may only display one of the following: predicted information on whether waste is input, predicted blowing amount of an induced blower, predicted blowing amount of a pressure blower, predicted blowing amount of an exhaust gas recirculation blower, predicted opening rate information of a blower damper, predicted chemical injection amount for reducing air pollutants, predicted chemical injection amount for improving incineration efficiency, or predicted pusher operation information. The blowing amount may include information on the predicted output frequency of an electric motor inverter connected to the blower. The user may input an input for controlling the incinerator based on the presented predicted operation control information to the operation control device (310). The operation control device (310) may control the incinerator based on the input received from the user.

[0233] The incinerator control system of the present disclosure generates predicted operation control information using an incinerator control reinforcement learning model, thereby assisting users in determining operation control information based on the predicted operation control information. Furthermore, the system can help incinerator users maintain the incinerator in optimal condition, even if they are not experts.

[0234] The incinerator control reinforcement learning model can be pre-generated by the server (100). The process of generating the incinerator control reinforcement learning model by the server (100) is described below. The process of generating the incinerator control reinforcement learning model can be performed by the server (100) or the data learning unit (110) of the operation control device (310).

[0235] FIG. 12 may be a drawing for explaining the operation of an incinerator control system according to one embodiment of the present disclosure.

[0236] The incinerator control system (300) can obtain learning operation status information and learning operation setting information. The incinerator control system (300) can obtain learning operation status information and learning operation setting information through the memory or sensor unit (320). The incinerator control system (300) can obtain learning difference information and learning trend information using at least one of the learning operation status information and the learning operation setting information. The incinerator control system (300) can obtain learning difference information and learning trend information using the method described in steps (530) and (540) of FIG. 5.

[0237] The process of the server (100) generating an incinerator control reinforcement learning model may be the same as the process of obtaining the highest reward in a Markov decision process. The server (100) may generate the incinerator control reinforcement learning model using at least one algorithm among Policy Optimization, Q-Learning, or Policy Optimization / Q-Learning Mixed. Policy Optimization may include Policy Gradient, A2C (Advantage Actor Critic), A3C (Asynchronous Advantage Actor-Critic), PPO (Proximal Policy Optimization), or TRPO (Trust region policy optimization). In addition, Q-Learning may include DQN (Deep Q Network), C51 (Categorical DQN), QR-DQN (Quantile Regression Deep Q Network), or HER (Hindsight Experience Replay). Policy Optimization / Q-Learning Mixed may include Deep Deterministic Policy Gradient (DDPG), Twin Delayed Deep Deterministic Policy Gradient (TD3), or Soft Actor Critic (SAC). The server (100) may typically utilize a Policy Optimization system, and may utilize PPO or SAC.

[0238] The incinerator control system (300) can obtain k-th learning state information (1210) using at least one of learning operation state information, learning operation setting information, learning difference information, and learning trend information. The k-th learning state information (1210) can indicate a state before the server (100) or the operation control device (310) takes action. The learning state information can include at least one of learning temperature information, the amount of learning nitrogen oxides, the amount of learning carbon monoxide, the amount of learning hydrogen chloride, the amount of learning sulfur oxides, the amount of learning dust, and the amount of waste input per hour.

[0239] The incinerator control system (300) can take one action among a plurality of candidate operation control information (1220) in the k-th learning state information (1210). The plurality of candidate operation control information (1220) can include one of information on whether waste is input, the weight of the input waste, the calorific value of the input waste, the calorific curve of the input waste, the blower's air volume, the opening rate of the blower damper, the amount of chemicals input to reduce air pollutants, the amount of chemicals input to improve incineration efficiency, and the operation cycle (speed) of the pusher (grate).

[0240] The incinerator control system (300) may perform a step of acquiring the kth learning operation control information (1221) from among a plurality of candidate operation control information (1220) available in the kth learning state information (1210) based on policy information. More specifically, the incinerator control system (300) may select one operation control information according to the probability of each of the plurality of candidate operation control information (1220). The incinerator control system (300) may derive the probability of each of the plurality of candidate operation control information (1220) based on a deep neural network. The probability of each of the plurality of candidate operation control information (1220) may be referred to as policy information of the incinerator control reinforcement learning model. The policy information is the probability of what action an action subject will take in a given environment. It may also mean a distribution of actions in the current incinerator state. Initially, the probability of each of the plurality of candidate operation control information (1220) may be randomly initialized. In the process of generating an incinerator control reinforcement learning model, the server (100) may update the probability for each of a plurality of candidate operation control information (1220) so that the incinerator control reinforcement learning model obtains the best reward. In other words, the policy information may be updated in the process of generating an incinerator control reinforcement learning model.

[0241] In addition, the incinerator control system (300) can select operation control information based on the above probability. The selected operation control information may be the kth learning operation control information (1221). The server (100) or the operation control device (310) may not necessarily select operation control information with a high probability, but may only select operation control information with a high probability. This is because, in order to train the incinerator control reinforcement learning model to balance exploration and exploitation, various actions must be attempted. In the process of generating the incinerator control reinforcement learning model, the server (100) may update the weights of the deep neural network using forward propagation and backpropagation to maximize the reward. The server (100) may initially randomly initialize the weights of the deep neural network, but the weights may be updated through forward propagation and backpropagation. As the weights are updated, the policy information of the incinerator control reinforcement learning model may be updated.

[0242] Based on the incinerator control reinforcement learning model, the server (100) can perform a step of acquiring k+1-th learning state information (1230) by reflecting the k-th learning operation control information (1221) to the k-th learning state information (1210). More specifically, the server (100) or the operation control device (310) can observe changed state information based on the k-th learning operation control information (1221) selected. The changed state information may be k+1-th learning state information (1230). The k+1-th learning state information (1230) may be information after a predicted time from the k-th learning state information (1210). The k+1-th learning state information (1230) may include operation state information after a predicted time, operation setting information after a predicted time, learning difference information after a predicted time, and learning trend information after a predicted time. For example, the learning status information may include learning temperature information, learning nitrogen oxide amount, learning carbon monoxide amount, learning hydrogen chloride amount, learning sulfur oxide amount, and learning dust amount.

[0243] The incinerator control system (300) can obtain the k+1-th learning state information (1230) based on a computer simulation. Alternatively, the server (100) or the operation control device (310) can obtain the k+1-th learning state information (1230) based on data obtained from an actual incinerator. Alternatively, the incinerator control system (300) can obtain the k+1-th learning state information (1230) by applying the k-th learning state information (1210) and the k-th learning operation control information (1221) to the operation state prediction machine learning model (1020). The incinerator control system (300) can use the k-th learning state information (1210) as operation state information, which is an input of the operation state prediction machine learning model (1020), and can use the k-th learning operation control information (1221) as operation setting information, which is an input of the operation state prediction machine learning model (1020).

[0244] The incinerator control system (300) may perform a step of determining a kth sub-reward of the kth learning operation control information for the kth learning state information by applying the k+1th learning state information to a reward function. The incinerator control system (300) may provide a kth sub-reward for the kth learning operation control information (1221) based on the k+1th learning state information (1230). The incinerator control system (300) may perform a step of generating an incinerator control reinforcement learning model by performing reinforcement learning so that the reward by the reward function (r) is maximized. The reward function (r) may be as follows.

[0245] r = r_temp + r_NOX + r_CO + r_HCL + r_SOX + r_Dust + r_waste

[0246] Here, r_temp may be a compensation value according to the learning temperature information. The learning temperature information may be temperature information at the outlet of the incinerator. r_NOX is a compensation value according to the amount of learning nitrogen oxide, r_CO is a compensation value according to the amount of learning carbon monoxide, r_HCL is a compensation value according to the amount of learning hydrogen chloride, r_SOX is a compensation value according to the amount of learning sulfur oxide, and r_Dust may be a compensation value according to the amount of learning dust. In addition, r_waste may be a compensation value according to the amount (weight) of waste input per learning time. The compensation function (r) is not limited to the above equation, and at least one of r_temp, r_NOX, r_CO, r_HCL, r_SOX, r_Dust, and r_waste may be omitted.

[0247] The reward determined by the reward function may be a sub-reward. That is, the incinerator control system (300) may determine the k-th sub-reward based on the above reward function. The server (100) may perform a step of determining the k-th reward based on the k-th sub-reward and the k+1-th sub-reward. That is, the server (100) may reflect not only the k-th sub-reward for the k-th learning operation control information (1221), but also the reward by the k+1-th learning operation control information in the k+1-th learning state information (1230) in the future to determine the final reward for the k-th learning operation control information (1221). This is because the result by the k-th learning operation control information (1221) will continue to have an influence in subsequent processes. The process of obtaining the k+1-th sub-reward is the same as the process of obtaining the k-th sub-reward, so a duplicate description will be omitted. The server (100) can determine the kth reward as the kth sub-reward + gamma * k+1th sub-reward + gamma^2 * k+2th sub-reward +...+ gamma^(a) * k+ath sub-reward. Here, gamma is a discount rate and can have a value greater than or equal to 0 and less than or equal to 1.

[0248] The incinerator control system (300) can perform a step of changing the policy information so that the kth reward is maximized. For example, the server (100) can change the selection probability for each of the plurality of candidate operation control information (1220) or change the weights of the neural network to change the policy information. The server (100) can perform a step of generating an incinerator control reinforcement learning model based on the policy information. The incinerator control system (300) can repeat the above process to increase the accuracy of the incinerator control reinforcement learning model.

[0249] The server (100) can store the incinerator control reinforcement learning model in memory. In addition, the server (100) can perform a step of transmitting the incinerator control reinforcement learning model to the operation control device (310) or another server.

[0250] Below, the process of determining the kth reward is described in detail together with Fig. 13.

[0251] FIG. 13 is a drawing for explaining the operation of an incinerator control system according to one embodiment of the present disclosure.

[0252] When the incinerator control system (300) performs the step of determining the kth sub-compensation, it can further perform the following process.

[0253] The incinerator control system (300) may perform a step of determining r_temp as a first compensation value when the learning temperature information included in the k+1-th learning state information (1230) is equal to or greater than the first threshold temperature and less than the second threshold temperature. Here, the learning temperature information may be temperature information of the incinerator outlet. In addition, the incinerator control system (300) may perform a step of determining r_temp as a second compensation value when the learning temperature information included in the k+1-th learning state information (1230) is equal to or greater than the third threshold temperature and less than the first threshold temperature, or equal to or greater than the second threshold temperature and less than the fourth threshold temperature. The incinerator control system (300) may perform a step of determining r_temp as a third compensation value when the learning temperature information included in the k+1-th learning state information (1230) is less than the third threshold temperature or greater than the fourth threshold temperature. The incinerator control system (300) can determine a compensation value according to the following conditions when the learning temperature information included in the k-th learning state information (1210) is less than the fifth threshold temperature or greater than the sixth threshold temperature. The server (100) can perform a step of determining r_temp as the fourth compensation value when the learning temperature information included in the k+1-th learning state information (1230) is greater than or equal to the seventh threshold temperature and less than the third threshold temperature, or greater than or equal to the fourth threshold temperature and less than the eighth threshold temperature.

[0254] The first compensation value, the second compensation value, the third compensation value, and the fourth compensation value may be predetermined values. The first compensation value, the second compensation value, and the fourth compensation value may be positive numbers, and the third compensation value may be negative. The first compensation value may be greater than the absolute value of the third compensation value, and the absolute value of the third compensation value may be greater than the second compensation value. According to one embodiment of the present disclosure, the fourth compensation value may be less than the second compensation value. However, the present disclosure is not limited thereto, and the fourth compensation value may be greater than the first compensation value.

[0255] The first to eighth critical temperatures may be predetermined temperatures. The relationship of the fifth critical temperature < seventh critical temperature < third critical temperature < first critical temperature < second critical temperature < fourth critical temperature < eighth critical temperature < sixth critical temperature may be satisfied. Additionally, the seventh critical temperature may be less than or equal to the third critical temperature. Furthermore, the fourth critical temperature may be less than or equal to the eighth critical temperature. The first critical temperature may be 900 degrees or more and 1000 degrees or less. The second critical temperature may be 1000 degrees or more and 1100 degrees or less. The third critical temperature may be 850 degrees or more and 950 degrees or less. The fourth critical temperature may be 1050 degrees or more and 1150 degrees or less. However, it is not limited thereto, and the values ​​of the critical temperatures may be changed by input from an administrator depending on the situation of the incinerator.

[0256] When the incinerator control system (300) performs the step of determining the kth sub-compensation, it can further perform the following process.

[0257] The incinerator control system (300) may perform a step of determining r_NOX as a fifth compensation value when the amount of learned nitrogen oxide included in the k+1-th learning state information is less than the amount of a predetermined reference nitrogen oxide. The predetermined reference nitrogen oxide may be an average of the amount of nitrogen oxide measured over a predetermined period of time. As described above, the amount of the material of the present disclosure may mean the mass of the material, or the molar concentration or mass concentration of the material in the whole.

[0258] The incinerator control system (300) may perform a step of determining r_NOX as a sixth compensation value when the amount of learned nitrogen oxide included in the k+1-th learning state information (1230) is greater than or equal to the amount of reference nitrogen oxide and less than a predetermined amount of allowable nitrogen oxide. The amount of allowable nitrogen oxide may be a predetermined value. For example, the amount of allowable nitrogen oxide may have a value of 30.0 ppm or more and 42.5 ppm or less. The incinerator control system (300) may perform a step of determining r_NOX as a seventh compensation value when the amount of learned nitrogen oxide included in the k+1-th learning state information is greater than or equal to the amount of allowable nitrogen oxide.

[0259] The fifth to seventh compensation values ​​may be predetermined values. The fifth compensation value may be positive, and the sixth and seventh compensation values ​​may be negative. In addition, the absolute values ​​of the fifth and sixth compensation values ​​may be less than the absolute value of the seventh compensation value. In addition, the amount of reference nitrogen oxides may be determined to be less than or equal to the amount of allowable nitrogen oxides.

[0260] When the incinerator control system (300) performs the step of determining the kth sub-compensation, it can further perform the following process.

[0261] The incinerator control system (300) may perform a step of determining r_CO as the eighth compensation value when the amount of learning carbon monoxide included in the k+1 learning state information (1230) is less than a predetermined reference carbon monoxide amount. The predetermined reference carbon monoxide amount may be an average of the amounts of carbon monoxide measured over a predetermined period of time. As described above, the amount of the material of the present disclosure may mean the mass of the material, or the molar concentration or mass concentration of the material in the whole.

[0262] The incinerator control system (300) may perform a step of determining r_CO as a ninth compensation value when the amount of learned carbon monoxide included in the k+1-th learning state information (1230) is greater than or equal to the amount of reference carbon monoxide and less than or equal to the amount of predetermined allowable carbon monoxide. The amount of allowable carbon monoxide may be a predetermined value. For example, the amount of allowable carbon monoxide may have a value of 35 ppm or more and 45 ppm or less. The server (100) may perform a step of determining r_CO as a tenth compensation value when the amount of learned carbon monoxide included in the k+1-th learning state information (1230) is greater than or equal to the amount of allowable carbon monoxide.

[0263] The eighth to tenth compensation values ​​may be predetermined values. The eighth compensation value may be a positive number, and the ninth and tenth compensation values ​​may be negative numbers. The absolute values ​​of the eighth and ninth compensation values ​​may be less than the absolute value of the tenth compensation value. The reference carbon monoxide amount may be determined to be less than or equal to the allowable carbon monoxide amount.

[0264] When the incinerator control system (300) performs the step of determining the kth sub-compensation, it can further perform the following process.

[0265] The incinerator control system (300) may perform a step of determining r_HCL as an 11th compensation value when the amount of learning hydrogen chloride included in the k+1 learning state information (1230) is less than a predetermined reference hydrogen chloride amount. The predetermined reference hydrogen chloride amount may be an average of the amounts of hydrogen chloride measured over a predetermined period of time. As described above, the amount of the material of the present disclosure may mean the mass of the material, or the molar concentration or mass concentration of the material in the whole.

[0266] The incinerator control system (300) may determine a step of determining r_HCL as a 12th compensation value when the amount of learning hydrogen chloride included in the k+1 learning state information is greater than or equal to the amount of reference hydrogen chloride and less than the amount of predetermined allowable hydrogen chloride. The amount of allowable hydrogen chloride may be a predetermined value. For example, the amount of allowable hydrogen chloride may have a value of 8.0 ppm or more and 9.6 ppm or less.

[0267] The incinerator control system (300) can perform a step of determining r_HCL as the 13th compensation value when the amount of learning hydrogen chloride included in the k+1 learning state information is greater than the amount of allowable hydrogen chloride.

[0268] The eleventh to thirteenth compensation values ​​may be predetermined values. The eleventh compensation value may be positive, and the twelfth and thirteenth compensation values ​​may be negative. The absolute values ​​of the eleventh and twelfth compensation values ​​may be less than the absolute value of the thirteenth compensation value. Additionally, the amount of reference hydrogen chloride may be less than or equal to the amount of allowable hydrogen chloride.

[0269] When the incinerator control system (300) performs the step of determining the kth sub-compensation, it can further perform the following process.

[0270] The incinerator control system (300) may perform a step of determining r_SOX as a 14th compensation value when the amount of learned sulfur oxides included in the k+1 learning state information (1230) is less than a predetermined reference sulfur oxide amount. The predetermined reference sulfur oxide amount may be an average of the amounts of sulfur oxides measured over a predetermined period of time. As described above, the amount of the material of the present disclosure may mean the mass of the material, or the molar concentration or mass concentration of the material in the entirety.

[0271] The incinerator control system (300) may perform a step of determining r_SOX as a 15th compensation value when the amount of learned sulfur oxides included in the k+1 learning state information (1230) is greater than or equal to the amount of reference sulfur oxides and less than the amount of predetermined allowable sulfur oxides. For example, the amount of allowable sulfur oxides may have a value of 15 ppm or more and 16 ppm or less.

[0272] The incinerator control system (300) may perform a step of determining r_SOX as a 16th compensation value when the amount of learned sulfur oxides included in the k+1 learning state information (1230) is greater than or equal to the amount of allowable sulfur oxides. The 14th compensation value to the 16th compensation value may be predetermined values. The 14th compensation value may be a positive number, and the 15th compensation value and the 16th compensation value may be negative numbers. The absolute values ​​of the 14th compensation value and the 15th compensation value may be less than the absolute value of the 16th compensation value. The amount of reference sulfur oxides may be less than or equal to the amount of allowable sulfur oxides.

[0273] When the incinerator control system (300) performs the step of determining the kth sub-compensation, it can further perform the following process.

[0274] The incinerator control system (300) may perform a step of determining r_Dust as a 17th compensation value when the amount of learning dust included in the k+1 learning state information (1230) is less than a predetermined reference dust amount. The predetermined reference dust amount may be an average of the amounts of dust measured over a predetermined period of time. As described above, the amount of the material of the present disclosure may mean the mass of the material, or the molar concentration or mass concentration of the material in the entirety.

[0275] The incinerator control system (300) may perform a step of determining r_Dust as the 18th compensation value when the amount of learning dust included in the k+1 learning state information is greater than or equal to the amount of reference dust and less than the amount of predetermined allowable dust. For example, the amount of allowable dust may have a value of 10 mg / Sm^3 or more and 12 mg / Sm^3 or less.

[0276] The incinerator control system (300) can perform a step of determining r_Dust as the 19th compensation value when the amount of learning dust included in the k+1 learning state information is greater than the amount of allowable dust.

[0277] The seventeenth to nineteenth compensation values ​​may be predetermined values. The seventeenth compensation value may be positive, and the eighteenth and nineteenth compensation values ​​may be negative. The absolute values ​​of the seventeenth and eighteenth compensation values ​​may be less than the absolute value of the nineteenth compensation value. Additionally, the amount of reference dust may be less than or equal to the amount of allowable dust.

[0278] When the incinerator control system (300) performs the step of determining the kth sub-compensation, it can further perform the following process.

[0279] The incinerator control system (300) can determine a reward value (r_waste) based on the amount of learning waste included in the k+1 learning state information (1230) and the amount of a predetermined reference waste. In addition, the incinerator control system (300) can assign a penalty point as the difference between the amount of learning waste and the amount of the predetermined reference waste increases. For example, the incinerator control system (300) can determine r_waste by multiplying the absolute value of the difference between the amount of input waste per unit time and the amount of reference waste by a predetermined constant.

[0280] r_waste = - a * |Amount of reference waste - Amount of input waste per unit time|

[0281] Here, the predetermined constant (a) can be a positive real number. The amount of waste can refer to the mass, weight, volume, or density of the material. In addition, the amount of waste input per unit time can refer to the mass information of the waste per unit time. The mass information of the waste can be included in the operating status information. Since the mass information of the waste has already been explained, a redundant explanation will be omitted. The amount of waste can refer to the mass, weight, volume, or density per hour. The amount of reference waste can be a predetermined value. For example, the amount of reference waste can be 5 tons or less per hour or 4 tons or more per hour. However, this is not limited to this, and the amount of reference waste can be determined differently depending on the incinerator. In addition, the amount of reference waste can be changed based on user input.

[0282] The server (100) can determine the kth sub-reward by determining r_temp, r_NOX, r_CO, r_HCL, r_SOX, r_Dust, and r_waste as described above. In addition, the server (100) can determine the kth reward by using the kth sub-reward to the k+ath sub-reward. The server (100) can generate an incinerator control reinforcement learning model by updating policy information so that the kth reward is maximized.

[0283] The incinerator control system of the present disclosure can generate operation control information predicted by an incinerator control reinforcement learning model. Furthermore, the system can assist a user in determining operation control information based on the predicted operation control information. Furthermore, the incinerator control system can automatically control the incinerator based on the predicted operation control information based on the incinerator control reinforcement learning model. Therefore, the incinerator control system can help maintain the incinerator in optimal condition even for non-expert users.

[0284] We have discussed various embodiments so far. Those skilled in the art will appreciate that the present invention can be implemented in modified forms without departing from its essential characteristics. Therefore, the disclosed embodiments should be considered illustrative rather than restrictive. The scope of the present invention is set forth in the claims, not the foregoing description, and all differences within the scope equivalent thereto should be construed as being encompassed by the present invention.

[0285] Meanwhile, the embodiments of the present invention described above can be written as a program that can be executed on a computer, and can be implemented in a general-purpose digital computer that runs the program using a computer-readable recording medium. The computer-readable recording medium includes storage media such as magnetic storage media (e.g., ROM, floppy disk, hard disk, etc.) and optical reading media (e.g., CD-ROM, DVD, etc.).

Claims

1. In the operation method of an incinerator control system including a sensor unit, an operation control device and a server, and controlling an incinerator using machine learning, A step in which the above driving control device acquires driving status information for a predetermined analysis time from the sensor unit; A step in which the above driving control device acquires driving setting information for the analysis time from the memory; A step of obtaining difference information between the current time and the previous time based on the driving status information and the driving setting information; A step of obtaining trend information during the analysis time based on the driving status information and the driving setting information; and An operating method of an incinerator control system, comprising a step of obtaining predicted operation control information by applying at least one of the operation status information, the operation setting information, the difference information, and the trend information to a driver simulation machine learning model received from the server.

2. In paragraph 1, The above driving status information is: Contains at least one of the following: mass information of waste, supply flow information, blower flow information, pressure information, storage tank information, temperature information, humidity information, and exhaust gas information; The mass information of the above waste includes at least one of the waste input time of the waste currently being incinerated in the incinerator, the weight of the input waste, the predetermined calorific value of the input waste, the predetermined calorific curve of the input waste, and the mass information of the waste input into the incinerator. The above supply flow information relates to the amount of at least one of urea water, ammonia water, ammonia gas, dilution water, slaked lime, magnesium hydroxide, water injection amount in the incinerator, deaerator feed water, cooling water feed water, economizer feed water, steam flow or boiler feed water supplied to the incinerator. The above blower flow rate information indicates the blower flow rate of at least one of a pressure blower, an induced blower, or an exhaust gas recirculator, The above pressure information relates to at least one of the pressure within the incinerator, the pressure within the boiler, the pressure within the cooling water pipe, the pressure within the boiler water pipe, the pressure within the ammonia tank, the pressure within the boiler drum, or the steam pressure. The information about the above storage tank relates to the degree of filling of at least one of a boiler feed water tank, a process water storage tank, a caustic soda feed tank, a slaked lime slurry storage tank, a urea water storage tank, an ammonia storage tank, a diesel storage tank, a cleaning water storage tank, a wastewater storage tank, a process water storage tank, a city water storage tank, or a magnesium hydroxide tank, The above temperature information is inside the incinerator, incinerator outlet, upper part of the incinerator drying stage, boiler inlet, boiler outlet, selective non-catalytic reduction device inlet, selective non-catalytic reduction device outlet, semi-dry washing tower inlet, semi-dry reaction tower outlet, centrifugal force collection facility (cyclone) inlet, centrifugal force collection facility (cyclone) outlet, filtration collection facility (bag filter) inlet, filtration collection facility (bag filter) outlet, washing tower inlet, washing tower outlet, selective catalytic reduction device inlet, selective catalytic reduction device outlet, stack discharge, deaerator inlet, deaerator outlet, boiler feedwater temperature, steam temperature, economizer inlet, economizer outlet, service water feed tank, process water storage tank, caustic soda supply tank, slaked lime slurry storage tank, urea water storage tank, ammonia storage tank, diesel storage tank, washing water storage tank, wastewater storage tank, process water storage tank, city water storage tank, or magnesium hydroxide tank, atmosphere, waste storage Relates to at least one temperature among the warehouse, hopper, primary air supply, secondary air supply, or exhaust gas recirculation inlet or exhaust gas recirculation outlet, and the ambient temperature of the incinerator facility; The above humidity information is provided at the inside of the incinerator, the incinerator outlet, the upper part of the incinerator drying stage, the boiler inlet, the boiler outlet, the selective non-catalytic reduction device inlet, the selective non-catalytic reduction device outlet, the semi-dry washing tower inlet, the semi-dry reaction tower outlet, the centrifugal dust collector (cyclone) inlet, the centrifugal dust collector (cyclone) outlet, the filtration dust collector (bag filter) inlet, the filtration dust collector (bag filter) outlet, the washing tower inlet, the washing tower outlet, the selective catalytic reduction device inlet, the selective catalytic reduction device outlet, the stack discharge, the water supply tank, the process water storage tank, the caustic soda supply tank, the slaked lime slurry storage tank, the urea solution storage tank, the ammonia storage tank, the diesel storage tank, the washing water storage tank, the wastewater storage tank, the process water storage tank, the city water storage tank, or the magnesium hydroxide tank, the atmosphere, the warehouse, the hopper section, the primary air supply, the secondary air supply, or the exhaust gas recirculation inlet or At least one of the humidity levels at the exhaust gas recirculation outlet and the incinerator facility is related to humidity, A method of operating an incinerator control system, wherein the information on the exhaust gas includes information on the amount of at least one of hydrogen chloride (HCL), nitrogen oxides (NOX), sulfur oxides (SOX), dust, carbon monoxide (CO), or oxygen (O2) emitted from the incinerator, and information on at least one of temperature, pressure, humidity, and flow rate.

3. In paragraph 1, The above operation setting information includes at least one of blower air volume information, pusher information, blower damper opening rate information, chemical injection amount information for reducing air pollutants, chemical injection amount information for improving incineration efficiency, or waste injection setting information. The above blower airflow information indicates the airflow of a pressurized blower, an induced blower, or an exhaust gas recirculation blower, and the airflow also indicates the output frequency of the electric motor inverter connected to the blower. The information about the above pusher indicates the operating cycle or the number of operations of the pusher included in the ramp pusher, the drying stage stocker, the combustion stage stocker, or the post-combustion stage stocker, An operating method of an incinerator control system, wherein the waste input setting information includes at least one of a waste input time, an input waste weight, and a predetermined heating curve of the input waste.

4. In paragraph 1, The above predicted driving control information is, Includes one of the following: information on whether the waste is input, the predicted blower volume of the induced blower, the predicted blower volume of the pressurized blower, the predicted exhaust gas recirculation blower volume, the predicted blower damper opening rate, the predicted chemical input amount to reduce air pollutants, the predicted chemical input amount to improve incineration efficiency, or the predicted pusher operation information. The above predicted waste input information indicates whether waste will be input or not. The predicted blowing volume of the above-mentioned induced blower includes information on the predicted output frequency of the electric motor inverter connected to the induced blower, The predicted blowing volume of the above-mentioned pressurized blower includes information on the predicted output frequency of the electric motor inverter connected to the above-mentioned pressurized blower, The predicted blowing volume of the exhaust gas recirculation blower includes information on the predicted output frequency of the electric motor inverter connected to the exhaust gas recirculation blower. The above predicted opening rate information of the blower damper indicates the predicted opening rate of a damper installed at least in one of the outlets of the induced blower, the pressurized blower, and the exhaust gas recirculation blower. The predicted chemical input amount information for reducing the above air pollutants includes at least one of the type of chemical input, the timing of input, and the amount of input. The predicted chemical input amount information for improving the above incineration efficiency includes at least one of the type of chemical input, the timing of input, and the amount of input. An operating method of an incinerator control system, wherein the predicted operation information of the pusher includes at least one of a cycle in which the pusher pushes waste and whether the pusher is operated (on / off).

5. In paragraph 4, A step in which the operation control device controls the incinerator based on the predicted operation control information; and An operating method of an incinerator control system further comprising a step of displaying at least one of the predicted information on whether or not waste is input, the predicted information on the blowing volume of the induced blower, the predicted information on the blowing volume of the pressurized blower, the predicted information on the blowing volume of the exhaust gas recirculation blower, the predicted information on the opening rate of the blower damper, or the predicted information on the operation of the pusher.

6. In paragraph 5, The above predicted driving control information indicates whether the predicted waste will be input or not. The step of controlling the incinerator by the operation control device based on the predicted operation control information is as follows: If the predicted waste input information indicates the input of waste, the operation control device performs a step of inputting waste; and An operating method of an incinerator control system including a step of not inputting waste when the predicted waste input information indicates that waste is not input.

7. In paragraph 1, The step of obtaining the above differential information is: A step of acquiring multiple learning driving status information and multiple driving setting information at a basic sampling cycle; A step of obtaining the driving status information of the current time, the driving setting information of the current time, the driving status information of the previous time, and the driving setting information of the previous time according to the system cycle based on the plurality of learning driving status information and the plurality of driving setting information; and It includes a step of obtaining the difference information by differentiating one of the driving status information of the previous time and the driving setting information of the previous time from one of the driving status information of the current time and the driving setting information of the current time, The difference between the current time and the previous time is the operating method of the incinerator control system, which is the system cycle.

8. In paragraph 1, The above trend information is an operation method of an incinerator control system including change amount information and change direction information of one of the above operation status information and the above operation setting information for each system cycle during the above analysis time.

9. In paragraph 8, The steps of obtaining the above trend information are: A step of acquiring multiple learning driving status information and multiple driving setting information at a basic sampling cycle; A step of obtaining driving status information at an n-th time according to a system cycle, driving setting information at the n-th time, driving status information at an n-1-th time, and driving setting information at the n-1-th time based on the plurality of learning driving status information and the plurality of driving setting information; A step of obtaining a plurality of differential information by differentiating one of the driving status information at the n-1th time and the driving setting information at the n-1th time from one of the driving status information at the n-th time and the driving setting information at the n-th time; and Including a step of obtaining trend information of the plurality of differential information, The difference between the above n-th time and the above n-1-th time is the system period, An operating method of an incinerator control system in which the sizes of the plurality of difference information correspond to the change amount information and the signs of the plurality of difference information correspond to the change direction information.

10. In paragraph 1, A step in which the server acquires a plurality of learning driving status information, a plurality of learning driving setting information, and a plurality of driving control information from a pre-collected learning database; A step in which the server acquires a plurality of past difference information and a plurality of past trend information based on the plurality of learning driving status information and the plurality of learning driving setting information; A step in which the server generates the driver simulation machine learning model by machine learning the causal relationship of the plurality of driving control information with respect to at least one of the plurality of learning driving status information, the plurality of learning driving setting information, the plurality of past difference information, and the plurality of past trend information; and An operating method of an incinerator control system, comprising a step of transmitting the above driver simulation machine learning model to the above driver control device.

11. In paragraph 10, The step of acquiring the plurality of learning driving status information, the plurality of learning driving setting information, and the plurality of driving control information is as follows: A step of acquiring candidate learning driving status information, candidate learning driving setting information, and candidate driving control information; A step of acquiring driving status information after a predetermined time corresponding to the candidate learning driving status information, the candidate learning driving setting information, and the candidate driving control information; A step of obtaining proficiency information based on driving status information after the above-determined time; and An operating method of an incinerator control system, comprising a step of determining the candidate operation control information as one of the plurality of operation control information when the above skill information is greater than or equal to a predetermined threshold compensation value.