Fire grate period control method and device, electronic equipment and computer storage medium

Abstract

The invention discloses a grate cycle control method and device, electronic equipment and a computer storage medium. The grate cycle control method specifically comprises the steps that parameter sequence values, the current flow value and the flow set value of main steam flow, the speed set value of a pusher, a hearth material layer pressure difference value, a pressure difference control variable sequence value and a material layer thickness calculated value are obtained; calculating a flow prediction value based on the parameter sequence value and the current flow value; calculating a flow deviation based on the flow set value and the flow predicted value; calculating a first target fire grate period based on the pusher speed given value; extracting a material layer thickness characteristic value based on the pressure difference control variable sequence value and the hearth material layer pressure difference value; calculating a second target fire grate period based on the material layer thickness characteristic value; calculating a third target fire grate period based on the material layer thickness calculation value; based on the flow prediction value and the target fire grate period, a target fire grate period given value is calculated; and based on the flow deviation and the target fire grate period given value, a target period is obtained and adopted to control the target fire grate.

Description

Technical Field

[0001] This disclosure belongs to the field of automatic combustion control technology, and in particular relates to a grate cycle control method and device, electronic equipment, and computer-readable storage medium. Background Technology

[0002] ACC (Automatic Combustion Control) is a core technology for the operation of incinerators, biomass boilers, and drying furnaces. It is mainly used to achieve automated and intelligent control of the combustion process, ensuring stable combustion, compliance with environmental standards, and economical operation. Its core principle is to optimize combustion efficiency by monitoring and adjusting parameters such as grate speed, air volume, and feed rate in real time.

[0003] Conventional ACC control methods establish a combustion chamber thermodynamic model based on the mass conservation equation and the energy conservation equation. However, due to large variations in composition, the mechanism model established based on the thermodynamic model is not applicable to actual production conditions. Summary of the Invention

[0004] To address the shortcomings of the prior art, this disclosure provides a grate cycle control method and apparatus, electronic equipment, and computer-readable storage medium.

[0005] To achieve the above objectives, this disclosure employs the following technical solution: The first aspect of this disclosure provides a grate cycle control method applied to an automatic combustion control system. The method includes: acquiring parameter sequence values ​​of control parameters related to the main steam flow rate during a historical period of the automatic combustion control system; the current flow rate value of the main steam flow rate; the flow rate setpoint value of the main steam flow rate; the pusher speed setpoint value; the furnace bed pressure difference value; and a sequence value of pressure difference control variables related to the furnace bed pressure difference height during a historical period, as well as a calculated bed thickness value related to the target grate bed thickness; calculating a predicted flow rate value of the main steam flow rate based on the parameter sequence values ​​and the current flow rate value; and calculating a predicted flow rate value of the main steam flow rate based on the flow rate setpoint and the current flow rate value. The flow rate is predicted, and the flow deviation is calculated. Based on the feeder speed setpoint, the first target grate cycle is calculated. Based on the differential pressure control variable sequence value and the furnace material layer differential pressure value, the material layer thickness characteristic value is extracted. Based on the material layer thickness characteristic value, the second target grate cycle is calculated. Based on the calculated material layer thickness value, the third target grate cycle is calculated. Based on the flow rate prediction value, the first target grate cycle, the second target grate cycle, and the third target grate cycle, the target grate cycle setpoint is calculated. Based on the flow rate deviation, the target grate cycle setpoint is dynamically corrected to obtain and use the target grate cycle to control the target grate.

[0006] A second aspect of this disclosure provides a grate cycle control device applied to an automatic combustion control system. The device includes: an acquisition unit configured to acquire parameter sequence values ​​of control parameters related to the main steam flow rate during a historical period of the automatic combustion control system parameter sequence values; the current flow rate value of the main steam flow rate; the flow rate setpoint value of the main steam flow rate; the pusher speed setpoint value; the furnace bed pressure difference value; and a pressure difference control variable sequence value related to the furnace bed pressure difference height during a historical period, as well as a calculated bed thickness value related to the target grate bed thickness; a prediction calculation unit configured to calculate a predicted flow rate value of the main steam flow rate based on the parameter sequence values ​​and the current flow rate value; and a deviation calculation unit configured to calculate a flow deviation based on the flow rate setpoint value and the predicted flow rate value; and a first cycle. The calculation unit is configured to calculate the first target grate cycle based on the feeder speed setpoint; the extraction unit is configured to extract the material layer thickness feature value based on the differential pressure control variable sequence value and the furnace material layer differential pressure value; the second cycle calculation unit is configured to calculate the second target grate cycle based on the material layer thickness feature value; the third cycle calculation unit is configured to calculate the third target grate cycle based on the calculated material layer thickness value; the setpoint calculation unit is configured to calculate the target grate cycle setpoint based on the flow prediction value, the first target grate cycle, the second target grate cycle, and the third target grate cycle; and the control unit is configured to dynamically correct the target grate cycle setpoint based on the flow deviation, and obtain and use the target grate cycle to control the target grate.

[0007] A third aspect of this disclosure provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a method as described in any implementation of the first aspect.

[0008] The fourth aspect of this disclosure provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method described in any implementation of the first aspect.

[0009] Compared with the prior art, the beneficial effects of this disclosure are as follows: Based on the control parameter sequence value and the current flow value, the flow prediction value is predicted and input into the control loop that controls the speed of the pusher, which can effectively improve the response speed of the automatic combustion control system. The flow deviation calculated by the flow prediction value can compensate for unnecessary fluctuations in the output of the automatic combustion control system in advance, thereby solving the overshoot problem caused by response lag, so that the control of the target grate can be closer to the actual production situation. Attached Figure Description

[0010] Figure 1This is a flowchart of an embodiment of the grate cycle control method according to the present disclosure; Figure 2 This is a schematic diagram of a structure of the grate cycle control method disclosed in this invention; Figure 3 This is another structural schematic diagram of the grate cycle control method disclosed in this paper; Figure 4 This is a schematic diagram of one embodiment of the grate cycle control device disclosed herein; Figure 5 This is a block diagram of an electronic device used to implement the grate cycle control method of the embodiments of this disclosure. Detailed Implementation

[0011] The present disclosure is further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the disclosure. Furthermore, it should be understood that after reading the teachings of this disclosure, those skilled in the art can make various alterations or modifications to this disclosure, and these equivalent forms also fall within the scope defined by this application.

[0012] The automatic combustion control system includes: feeding equipment, grate system, blower system, slag removal system, monitoring equipment, and a control system comprising multiple control modules. The feeding equipment includes a pusher for uniformly and stably feeding material into the incinerator; its speed can be adjusted according to combustion conditions. The control system includes: evaporation control module, material layer control module, furnace temperature control module, slag ignition rate control module, and oxygen control module. The evaporation control module ensures stable steam production by adjusting the combustion process, guaranteeing process requirements. The grate system includes: drying grate, combustion grate, and burnout grate located in different sections of the furnace. Segmented combustion is achieved by controlling the operating speed and air supply of each section. The blower system includes primary and secondary blowers. The primary blower supplies combustion air to the area under the grate, while the secondary blower supplies air to the secondary combustion chamber; both are controlled by frequency converters.

[0013] In the automatic combustion control system, the transport vehicle delivers the material to the incineration plant, unloads it into the pool, and then the material enters the drying chamber via a conveying assembly. Heating pipes within the drying chamber pre-dry the material. The bed control module calculates the bed thickness by measuring the air pressure difference between the top and bottom of the combustion grate, and then adjusts the speed of the pusher, drying grate, and combustion grate to stabilize the bed thickness. In the drying grate, the material is dried by high-temperature combustion air, radiant heat from the furnace side walls and top, and the residence time is approximately 30 minutes. After drying, the material enters the combustion section for vigorous combustion, with a residence time of approximately 30 minutes, during which 60%-80% of the total combustion air is supplied. In the burnout section, the fixed carbon and unburned portions are completely burned, with a residence time of approximately one hour. The system ensures that the slag's loss on ignition is reduced to 1%-2%. Air drawn from above the primary air intake by the primary air fan is heated by a steam-air preheater and then sent to the grates in the drying, combustion, and burnout sections, using variable frequency drive (VFD) control. Air drawn from above the secondary air intake by the secondary air fan is heated by a steam-air preheater and then supplied to the secondary combustion chamber, also using VFD control. During combustion, parameters such as primary air volume, secondary air volume, and feed rate are adjusted to maintain the furnace temperature within a suitable range, ensuring stable combustion. Furthermore, based on the oxygen content in the flue gas measured by the oxygen sensor, PID or fuzzy control is used to adjust the speed of the primary and secondary air fans, changing the primary and secondary air volumes to maintain the oxygen content in the flue gas at a reasonable level.

[0014] Current automatic combustion control systems establish a thermodynamic model of the combustion chamber based on the mass conservation equation and the energy conservation equation. They use real-time detected values ​​such as furnace temperature, flue gas oxygen content, and steam flow rate as input variables, and output execution commands such as grate speed and primary / secondary fan frequency through a PID controller. However, due to large composition variations, the mechanism model based on thermodynamics is not applicable to actual production conditions. Furthermore, there is a long lag period from sensor detection to actuator response, resulting in severe control overshoot and large system fluctuations.

[0015] To address the shortcomings of traditional technologies, this disclosure proposes a grate cycle control method for use in automatic combustion control systems. Figure 1 A flow chart 100 of an embodiment of the grate cycle control method according to the present disclosure is shown, the grate cycle control method comprising the following steps: Step 101: Obtain the parameter sequence values ​​of the control parameters related to the main steam flow rate, the current flow rate value of the main steam flow rate, the flow rate setpoint value of the main steam flow rate, the feeder speed setpoint value, the furnace material layer pressure difference value, the pressure difference control variable sequence values ​​related to the furnace material layer pressure difference height, and the material layer thickness calculation value related to the target grate material layer thickness in the historical period of the automatic combustion control system.

[0016] In this embodiment, the historical period is the historical time period of the waste automatic combustion control system. The duration of the historical period can be set according to development needs, such as a historical period of three months. The control parameters are the parameters that affect the main steam flow in the waste automatic combustion control system. These control parameters can be obtained by analyzing parameters through correlation algorithms (such as Pearson's algorithm and Spearman's algorithm), such as main steam pressure, furnace average temperature, and flue gas temperature at the outlet flue. The control parameter sequence value includes the value of at least one control parameter. The value of the control parameter can be obtained by sensors. The control parameter sequence value is a serialized representation of the control parameter value in the historical period.

[0017] In a specific example, the parameter sequence values ​​are obtained through the following steps: obtaining the historical parameter set of the automatic combustion control system; calculating the correlation coefficient between the historical parameter set and the main steam flow rate under different delays; and determining the parameter sequence values ​​based on the correlation coefficient.

[0018] In this optional implementation, the calculation of the correlation coefficient between the historical parameter set and the main steam flow rate under different delays includes: determining multiple delay times; specifically, these multiple delay times are different delay times, such as specific times like 0s, 5s, 10s, etc., or they can be identified by the number of sampling periods, where one sampling period can be 5s or 10s, etc. Determining the delay time means determining the delay interval and the number of delays. Multiple delay times constitute a continuous time; for example, if every 5s is a delay time and there are 20 delays, the entire delay process is 100s. The delay time characterizes the time when changes in the parameters of the automatic waste combustion control system cause significant changes in the parameters to be predicted in the automatic waste combustion control system (such as the predicted value of the main steam flow rate). Under each delay time, the correlation coefficient between each historical parameter in the historical parameter set and the main steam flow rate is calculated. Specifically, the Pearson algorithm or the Spearman algorithm can be used to calculate the correlation coefficient.

[0019] In this optional implementation, determining the parameter sequence values ​​based on the correlation coefficient includes: selecting at least one parameter with the highest correlation coefficient in the automatic waste combustion control system as the control parameter, and extracting the parameter values ​​of this control parameter from historical periods to obtain the parameter sequence values. In actual calculations, the control parameters related to the main steam flow include: steam drum pressure, evaporator inlet temperature (left), main feedwater pressure, evaporator inlet temperature (right), main feedwater flow rate, main steam pressure, first-stage desuperheating water flow rate, furnace average temperature, second-stage desuperheating water flow rate, outlet flue gas temperature (right), economizer inlet oxygen concentration, furnace flue gas temperature (left), furnace flue gas temperature (middle), furnace flue gas temperature (right), furnace outlet flue gas temperature (right), economizer outlet flue gas pressure (left), economizer outlet flue gas pressure (right), NOx concentration, furnace outlet flue gas temperature (left), economizer outlet oxygen concentration (right), pusher speed setpoint, and main steam flow rate. The parameter sequence values ​​of these control parameters can be values ​​collected by field sensors.

[0020] In this optional implementation, the target delay (or prediction step size k*) is defined as the delay corresponding to the maximum value of the correlation between the control parameter (actively changing parameter) and the controlled parameter (follower parameter). Specifically, it is the response time required for the controlled parameter to reach a steady state after a step change in the control parameter.

[0021] The target delay can be obtained through the step response method or the correlation delay algorithm (cross-correlation method). The step response method is a mature algorithm and will not be elaborated upon here. The core of the correlation delay algorithm lies in: first, determining the multi-step delay time range of the control parameter x and the controlled parameter y. The multi-step delay time range is the range of the delay step k (k is a non-negative integer; typically, the multi-step delay time range is limited to 0 to 2000. This non-negative limitation is because online prediction can only be based on historical data and cannot utilize future data). For each delay step k within this multi-step delay time range (each delay step corresponds to a fixed time interval, such as 5 seconds), a time-aligned control parameter sequence x(t) and a controlled parameter sequence y(t+k) (or equivalent form) need to be constructed. The construction of the sequence is essentially a sliding window process: a new sequence under each delay is obtained by removing the last value of the control parameter and the first value of the controlled parameter from the previous sequence and realigning the time points. For example, when k=0, x(t) and y(t) are aligned (no delay); when k=1, x(t) and y(t+1) are aligned (representing that the y value lags behind the x value by 1 sampling interval, i.e., 5 seconds); when k=2, x(t) and y(t+2) are aligned (lagging by 10 seconds), and so on. Subsequently, the Pearson correlation coefficient ρ(k) of each delay k sequence x(t) and y(t+k) is calculated, the maximum value among all |ρ(k)| is found, and the delay k corresponding to the maximum value is determined as the target delay k*, the mathematical expression of which is shown in Equation (1).

[0022] (1) The k* calculated by this method comprehensively reflects the transmission delay of the physical process and the response time of the measuring instruments.

[0023] This implementation provides a reliable method for obtaining control parameter sequence values ​​(or prediction step sizes) by calculating the correlation coefficients under different delays and determining the delay corresponding to the maximum value. In this embodiment, the main steam flow rate setpoint is a target value set by the user based on historical experience. The flow rate setpoint is determined according to the actual conditions of the combustion process and the furnace. This flow rate setpoint aims to maintain the main steam flow rate at an optimal level, ensuring combustion efficiency and stability.

[0024] In this embodiment, the furnace bed pressure difference refers to the pressure difference generated on both sides of the waste bed when airflow passes through it within the furnace. It is an important indicator of the resistance of the waste bed to airflow. In a waste incinerator, the magnitude of the furnace bed pressure difference reflects the accumulation of the waste bed. As the thickness of the waste bed increases, the resistance to airflow through the bed increases, leading to an increase in the furnace bed pressure difference. By monitoring changes in the furnace bed pressure difference, the trend of bed thickness changes can be determined. For example, if the furnace bed pressure difference gradually increases from an initial 1000 Pa to 1500 Pa, this may indicate that the bed thickness is increasing. A differential pressure transmitter (used to prevent the medium in the pipeline from directly entering the transmitter; the pressure-sensing diaphragm is connected to the transmitter by a capillary tube filled with fluid. It is used to measure the level, flow rate, and pressure of liquids, gases, or steam, and then converts them into a 4-20mA DC signal output) can measure the pressure difference of the furnace bed. A differential pressure transmitter generally has two probes, installed on both sides of the medium, to calculate the pressure difference between the two sides.

[0025] In this embodiment, the pusher speed setpoint is the speed value input to the pusher. This speed value can be used to control the speed of the pusher. The pusher speed setpoint can be determined by comprehensively considering various factors and control strategies. The specific calculation steps include: First, calculating the base speed through the preset main steam flow rate setpoint. The pusher base speed is closely related to the main steam flow rate setpoint. When the main steam flow rate setpoint changes, the baseline value of the fuel entering the furnace changes through the change in the pusher speed. Specifically, when the main steam flow rate setpoint increases, the pusher speed automatically increases; when the main steam flow rate setpoint decreases, the pusher speed automatically decreases. Assuming that the relationship between the main steam flow rate setpoint and the pusher base speed is linear, it can be expressed as shown in equation (2): The basic speed of the pusher = k × main steam flow rate setpoint + b (2) In formula (2), k and b are coefficients determined based on actual working conditions.

[0026] The second step involves adjusting the basic speed of the pusher based on the deviation between the measured value of the flue gas oxygen content and the preset value, thus obtaining the calculated speed of the pusher. When the measured value of the flue gas oxygen content is higher than the preset value, the pusher speed is reduced; when the measured value of the flue gas oxygen content is lower than the preset value, the pusher speed is increased. Assuming the compensation factor is Δv, the calculated speed of the pusher can be expressed as shown in equation (3): The calculated speed of the pusher = the basic speed of the pusher + Δv (3) The third step involves further correcting the pusher's calculated speed based on the deviation between the calculated material layer thickness and the preset material layer thickness setting, ultimately obtaining the pusher speed setting value. When the calculated material layer thickness is higher than the setting value, the pusher speed is reduced; when the calculated material layer thickness is lower than the setting value, the pusher speed is increased. Assuming the correction factor caused by the material layer thickness deviation is Δv′, the pusher speed setting value can be expressed as shown in equation (4): Pusher speed setting value = Pusher calculated speed + Δv′ (4) In this embodiment, the differential pressure control variable related to the furnace bed pressure difference height is a parameter in the automatic waste combustion control system that affects the furnace bed pressure difference height. This differential pressure control variable can be obtained by analyzing variables through correlation algorithms (such as Pearson's algorithm and Spearman's algorithm), such as primary air fan frequency, secondary air fan frequency, furnace negative pressure, drying grate primary air damper opening, combustion grate primary air damper opening, and burnout grate primary air damper opening. The differential pressure control variable sequence value includes the value of at least one differential pressure control variable. The value of the differential pressure control variable can be acquired by sensors. This differential pressure control variable sequence value is a serialized representation of the control variable values ​​in history.

[0027] In this embodiment, the grate cycle control method is a method for controlling the target grate cycle, wherein the target grate cycle is the time required for the target grate to complete one complete action cycle. For example, the movement cycle of the target grate is 80s, which requires 10s of movement and 70s of rest.

[0028] In this embodiment, the calculated material layer thickness is the thickness value that affects the thickness of the target grate in the automatic combustion control system. These parameters may not be obtained through correlation analysis, but rather selected based on process theory (mechanism-driven rather than data-driven). For example, when the target grate is a dry grate, the calculated material layer thickness related to the target grate material layer thickness is: the material layer thickness of the combustion grate; when the target grate is a combustion grate, the calculated material layer thickness related to the target grate material layer thickness is: the material layer thickness of the combustion grate and the material layer thickness of the burnout grate; when the target grate is a burnout grate, the calculated material layer thickness related to the target grate material layer thickness is: the material layer thickness of the burnout grate.

[0029] In this embodiment, the calculated value of the material layer thickness related to the target grate material layer thickness can be obtained by taking an image of the material layer in the target furnace and using an image segmentation algorithm to calculate the material layer in the image information.

[0030] Step 102: Calculate the predicted flow rate of the main steam flow rate based on the parameter sequence value and the current flow rate value.

[0031] In this embodiment, the parameter sequence value is a serialized representation of the control parameter values ​​in the historical period. Since the main steam flow rate has the highest correlation with itself, the control parameter includes the main steam flow rate. By arranging the parameter sequence value and the current flow rate value together and putting them into the pre-trained time series prediction model, the time series prediction model can refer to the control parameters other than the main steam flow rate in the parameter sequence value to predict the future value of the main steam flow rate and obtain the flow prediction value.

[0032] Optionally, step 102 above includes: extracting the main steam flow rate and the parameter values ​​of control parameters other than the main steam flow rate from the parameter sequence values; fitting the main steam flow rate curve based on the parameter values ​​of the main steam flow rate; fitting a development trend curve based on the parameter values ​​of control parameters other than the main steam flow rate; adjusting the main steam flow rate curve based on the development trend curve to obtain an adjustment curve; and obtaining the predicted flow rate value of the main steam flow rate based on the adjustment curve.

[0033] Step 103: Calculate the flow deviation based on the flow setpoint and the flow prediction.

[0034] In this embodiment, the flow rate deviation can be obtained by directly subtracting the flow rate prediction value from the flow rate setting value.

[0035] Optionally, step 103 includes: obtaining the parameter value of the main steam flow rate from the control parameter sequence value; determining the upper limit and lower limit of the parameter value of the main steam flow rate; subtracting the lower limit from the upper limit to obtain the limit difference; multiplying the limit difference by a preset coefficient to obtain the deviation limit; subtracting the flow prediction value from the flow set value to obtain the first difference; detecting whether the first difference is within the deviation limit; if the first difference is within the deviation limit, using the first difference as the flow deviation.

[0036] Step 104: Calculate the first target grate cycle based on the feeder speed setpoint.

[0037] In this embodiment, the feeder speed setpoint is input into the trained cycle prediction model to obtain the first target grate cycle output by the cycle prediction model. The cycle prediction model is trained using samples generated from the target grate cycle of the waste automatic combustion control system and the speed of the waste feeder. Figure 2 As shown, the first target grate cycle (not shown in the figure) can be obtained by using the feeder speed setpoint and the speed piecewise linear function F1(x).

[0038] Step 105: Extract the characteristic value of the material layer thickness based on the differential pressure control variable sequence value and the differential pressure value of the furnace material layer.

[0039] In this embodiment, the bed thickness feature value refers to the thickness characteristic of the fuel in the furnace of the automatic combustion control system. This feature value can be extracted using a mathematical model or machine learning algorithm. Assuming a machine learning model trained on historical data, it uses the sequence values ​​of pressure difference control variables (such as the waste feeding rate sequence, secondary air volume sequence, etc.) and the furnace bed pressure difference value as input features. The model predicts the bed thickness based on the complex relationships between these input features, providing the bed thickness feature value. For example, the model might find that when the waste feeding rate increases within a certain range, and the furnace bed pressure difference value also increases accordingly, the bed thickness will change according to a certain pattern. In this way, the model can output a bed thickness feature value, which can be a specific thickness value or an indicator representing the trend of thickness change.

[0040] Step 106: Calculate the second target grate cycle based on the material layer thickness characteristic value.

[0041] In this embodiment, the second target grate cycle is the grate cycle determined by the characteristic value of the material layer thickness. In the automatic waste combustion control system, if the material layer thickness is large, it is necessary to control the target grate cycle in order to maintain combustion efficiency and prevent waste accumulation.

[0042] In this embodiment, the second target grate cycle can be calculated based on the deviation between the characteristic value of the material layer thickness and the set target material layer thickness. Specifically, the steps include: comparing the extracted characteristic value of the material layer thickness with the preset target material layer thickness to calculate the material layer deviation value. The material layer deviation value is the difference between the characteristic value of the material layer thickness and the target material layer thickness. The material layer deviation value is input into a PID algorithm, which calculates the second target grate cycle based on the magnitude and trend of the deviation value. The preset target material layer thickness is set according to the combustion process and the actual conditions of the furnace. This target material layer thickness is the optimal thickness of the waste material layer that is desired to be maintained in the furnace, ensuring combustion efficiency and stability. Figure 2 As shown, the second target grate cycle (not shown in the figure) can be obtained by using the characteristic value of the material layer thickness and the material layer piecewise linear function F2(x).

[0043] Step 107: Calculate the third target grate cycle based on the calculated material layer thickness.

[0044] In this embodiment, the calculated material layer thickness is obtained by calculating the thickness of fuel in the target grate or a grate related to the target grate. The third target grate cycle is the grate cycle determined by the calculated material layer thickness. Since the calculated material layer thickness directly affects the thickness of the target grate, the target grate cycle can be directly calculated using the calculated material layer thickness. Figure 2As shown, the third target grate cycle (not shown in the figure) can be obtained by combining the calculated material layer thickness with the target thickness piecewise linear function F3(x). Step 107 above includes: obtaining the target thickness piecewise linear function that characterizes the relationship between the calculated material layer thickness and the target grate cycle; and obtaining the third target grate cycle based on the calculated material layer thickness and the target thickness piecewise linear function.

[0045] The target thickness piecewise linear function can be a piecewise linear function that establishes a linear relationship between input and output variables using big data technology. The process of obtaining the target thickness piecewise linear function is as follows: Calculated values ​​of the target grate thickness and the target grate cycle are collected from historical time periods. Based on the distribution of the calculated target grate thickness and the target grate cycle, a suitable function form is selected for fitting. For example, a linear function, a polynomial function, or an exponential function can be chosen to obtain the target thickness piecewise linear function.

[0046] Step 108: Calculate the target grate cycle setpoint based on the flow prediction value, the first target grate cycle, the second target grate cycle, and the third target grate cycle.

[0047] In this embodiment, the third target grate cycle can be determined by the correspondence between the predicted flow rate and the pre-calibrated flow rate. The first target grate cycle, the second target grate cycle, and the third target grate cycle are added together to obtain the pusher speed setpoint.

[0048] Step 109: Based on the flow deviation, dynamically correct the target grate cycle setpoint to obtain and use the target grate cycle to control the target grate.

[0049] In this embodiment, a fuzzy logic controller (FLC) can be used to obtain the target speed value of the pusher. A fuzzy logic controller is a control method based on fuzzy logic, suitable for handling processes with high uncertainty and fuzziness. The following are the specific implementation steps for obtaining the target speed value of the pusher using a fuzzy logic controller: The input variables of the fuzzy logic controller are: flow deviation ΔF; flow deviation change rate ΔF′, which is the rate of change of flow deviation, i.e., the derivative of ΔF. The output variable is: target periodic correction value ΔV: the target periodic correction value calculated based on the flow deviation and the flow deviation change rate.

[0050] The fuzzy logic controller divides input and output variables into fuzzy sets. For example, the flow deviation ΔF can be divided into fuzzy sets such as "negative large" (NB), "negative small" (NS), "zero" (Z), "positive small" (PS), and "positive large" (PB). Fuzzy rules are defined based on experience and expert knowledge. For example: if ΔF is "negative large" and ΔF′ is "negative large", then ΔV is "positive large"; if ΔF is "zero" and ΔF′ is "zero", then ΔV is "zero"; if ΔF is "positive large" and ΔF′ is "positive large", then ΔV is "negative large".

[0051] The fuzzy logic controller calculates the membership degree of fuzzy sets. Based on the actual value of the input variable, it calculates the membership degree of the variable in each fuzzy set. For example, if ΔF = 10 tons / hour, it calculates the membership degree of the variable in the fuzzy sets "negative large", "negative small", "zero", "positive small", and "positive large".

[0052] The fuzzy logic controller determines the fuzzy set of the output variables based on fuzzy rules and membership degrees. For example, according to the fuzzy rules mentioned above, if the membership degree of ΔF is higher in "positive large" and the membership degree of ΔF′ is higher in "positive large", then the membership degree of ΔV is higher in "negative large".

[0053] The fuzzy logic controller converts fuzzy outputs into specific numerical values. Common defuzzification methods include the maximum membership method and the centroid method. For example, the centroid method is used to calculate the specific value of ΔV; the defuzzified ΔV is added to the target grate period setpoint Vinital to obtain the target grate period Vtarget, i.e., Vtarget = Vinital + ΔV; the calculated target grate period Vtarget is used as a control signal and sent to the target grate drive system, which adjusts the actual period according to the received target period.

[0054] The grate cycle control method provided in the embodiments of this disclosure acquires the parameter sequence values ​​of control parameters related to the main steam flow rate, the current flow rate value of the main steam flow rate, the flow rate setpoint value of the main steam flow rate, the pusher speed setpoint value, the furnace bed pressure difference value, and the pressure difference control variable sequence values ​​related to the furnace bed pressure difference height and the calculated bed thickness value related to the target grate bed thickness during the historical period of the automatic combustion control system; calculates the predicted flow rate value of the main steam flow rate based on the parameter sequence values ​​and the current flow rate value; and calculates the flow rate based on the flow rate setpoint value and the predicted flow rate value. The process involves several steps: First, the target grate cycle is calculated based on the pusher speed setpoint. Second, the target grate cycle is calculated based on the differential pressure control variable sequence and the furnace bed differential pressure. Third, the target grate cycle setpoint is calculated based on the predicted flow rate, the first target grate cycle, the second target grate cycle, and the third target grate cycle. Finally, the target grate cycle setpoint is dynamically corrected based on the flow rate deviation to obtain and use the target grate cycle to control the target grate. This method, by predicting the flow rate based on the control parameter sequence and the current flow rate, and inputting the predicted flow rate into the control loop controlling the pusher speed, effectively improves the response speed of the automatic combustion control system. Calculating the flow rate deviation using the predicted flow rate can compensate for unnecessary fluctuations in the output of the automatic combustion control system, thus solving the overshoot problem caused by response lag. This allows for more precise control of the target grate, closely mirroring actual production conditions.

[0055] Optionally, the above-mentioned grate cycle control method further includes: acquiring flame images of grates related to the target grate (e.g., grates related to the drying grate are combustion grates, grates related to the combustion grate are the combustion grate itself, and grates related to the burnout grate are the burnout grate itself) using a camera (e.g., an infrared thermal imager); preprocessing, hole-filling, and edge detection of the acquired flame images; extracting flame features of the material pile area; then, using the furnace pushing platform size, camera resolution, and field of view information, calculating the actual size of a single pixel; and finally, processing the obtained material pile flame image based on the pixel size to obtain... The process involves: analyzing the flame characteristics of the combustion grate; calculating an additional target grate cycle based on these characteristics; and calculating the target grate cycle setpoint based on the predicted flow rate, the first target grate cycle, the second target grate cycle, and the third target grate cycle. This includes: obtaining a flow rate piecewise linear function representing the relationship between the main steam flow rate and the grate cycle; obtaining a fourth target grate cycle based on the predicted flow rate and the flow rate piecewise linear function; and weighted summing the first target grate cycle, the second target grate cycle, the third target grate cycle, the fourth target grate cycle, and the additional target grate cycle to obtain the target grate cycle setpoint. In this embodiment, a computer vision algorithm is used to extract the actual flame thickness of the combustion grate as a factor influencing the target grate cycle setpoint, thereby allowing for targeted adjustments to the target grate cycle and improving the accuracy of obtaining the target grate cycle.

[0056] In some optional implementations of this disclosure, the target grate is a combustion grate, and the calculated material layer thickness includes: the combustion grate thickness value and the burnout grate thickness value; the calculation of the third target grate cycle based on the calculated material layer thickness value includes: calculating the first sub-grate cycle based on the combustion grate thickness value; calculating the second sub-grate cycle based on the burnout grate thickness value; and obtaining the third target grate cycle based on the first sub-grate cycle and the second sub-grate cycle.

[0057] In this optional implementation, the first sub-grate cycle is the cycle value determined by the combustion grate in the third target grate cycle, and the second sub-grate cycle is the cycle value determined by the burnout grate in the third target grate cycle. The third target grate cycle is obtained by weighted summing of the first sub-grate cycle and the second sub-grate cycle.

[0058] In this optional implementation, the calculation of the first sub-grate period based on the combustion grate thickness value includes: obtaining a first thickness piecewise linear function characterizing the correspondence between the combustion grate thickness value and the combustion grate period, such as... Figure 3 As shown, F31(x) represents the first thickness piecewise linear function; based on the combustion grate thickness value and the first thickness piecewise linear function, the first sub-grate period is obtained.

[0059] In this optional implementation, the calculation of the second sub-grate period based on the burnt-out grate thickness value includes: obtaining a second thickness piecewise linear function characterizing the correspondence between the burnt-out grate thickness value and the combustion grate period, such as... Figure 3 As shown, F32(x) represents the first thickness piecewise linear function; based on the burnout grate thickness value and the second thickness piecewise linear function, the second sub-grate period is obtained.

[0060] In this optional implementation, both the first and second thickness piecewise linear functions can be piecewise linear functions that establish a linear relationship between input and output variables using big data technology. The process of obtaining the first or second thickness piecewise linear function is as follows: Multiple test grate thickness values ​​(the test grate can be a burning grate or a burnout grate; when establishing the first thickness piecewise linear function, the test grate is a burning grate; when establishing the second thickness piecewise linear function, the test grate is a burnout grate) and the burning grate period are collected from historical time periods. Based on the distribution of the test grate thickness values ​​and the burning grate period, a suitable function form is selected for fitting. For example, a linear function, a polynomial function, or an exponential function can be selected to obtain the first or second thickness piecewise linear function.

[0061] In this optional implementation, if the target grate is a combustion grate, then the target period of the target grate calculated by the grate period control method is the target period of the combustion grate. Figure 3 As shown. It should be noted that when calculating the target cycle of the combustion grate, the predicted flow rate, the feeder speed setpoint, and... The method for calculating the third target grate cycle provided by this optional implementation refers to the thickness values ​​of the burning grate and the burnout grate, respectively, to calculate the third target grate cycle, thereby improving the reliability and accuracy of obtaining the third target grate cycle.

[0062] In some optional implementations of this disclosure, the above-mentioned calculation of the main steam flow prediction value based on parameter sequence values ​​and flow values ​​includes: constructing feature parameter values ​​based on parameter sequence values ​​and flow values; inputting the feature parameter values ​​into a pre-trained main steam flow prediction model to obtain the main steam flow prediction value.

[0063] In this optional implementation, the construction of feature parameter values ​​based on parameter sequence values ​​and current flow rate values ​​includes: extracting parameter values ​​other than the main steam flow rate from the parameter sequence values ​​and using these parameter values ​​as input features; extracting the parameter value of the main steam flow rate and the current flow rate value from the parameter sequence values ​​and using them as target variables or output features; and combining the input features and target variables or output features into an input feature matrix. Assuming there are three control parameters (pusher speed setpoint, main steam flow rate, and secondary air volume) and a time step of 10, the shape of the input feature matrix for each sample is (10, 4), where 4 represents the three control parameters and one flow rate value.

[0064] In this optional implementation, the values ​​of each control parameter in the parameter sequence are collected every 5 seconds. Initialization requires 200 collections (1000 seconds) to form a 200*22 two-dimensional input feature matrix, which is then input into the main steam flow prediction model. The main steam flow prediction model outputs a one-dimensional array of length 33. The first element of the array is the predicted value after 5 seconds, and the 33rd element is the flow prediction value after 165 seconds. Afterward, the input variables are collected every 5 seconds to fill the last row of the input matrix, while the first row (the oldest row) is deleted, ensuring the input matrix always maintains a size of 200*22. The next cycle of inference is then performed, and this process is repeated.

[0065] In this optional implementation, the main steam flow prediction model can employ a Long Short-Term Memory (LSTM) network. LSTM is a special type of recurrent neural network capable of learning long-term dependencies. It controls the flow of information through a gating mechanism, avoiding the gradient vanishing problem of traditional recurrent neural networks. The LSTM model includes an input layer, an LSTM layer, a fully connected layer, and an output layer. Each LSTM layer contains an input gate, a forget gate, and an output gate, with each gate controlling the inflow and outflow of information. LSTM is suitable for time series data that requires learning long-term dependencies.

[0066] The method for calculating the predicted flow rate of the main steam flow provided by this optional implementation improves the reliability and accuracy of the predicted flow rate by predicting the characteristic parameter values ​​formed by the control parameter sequence and the current flow rate value through the main steam flow prediction model.

[0067] In some optional implementations of this disclosure, the above-mentioned calculation of the first target grate cycle based on the feeder speed setpoint includes: obtaining a velocity piecewise linear function characterizing the relationship between the target grate cycle and the feeder speed; and obtaining the first target grate cycle based on the feeder speed setpoint and the velocity piecewise linear function.

[0068] In this optional implementation, the speed piecewise linear function is a function that establishes a linear relationship between the input and output variables using big data technology. The process of obtaining the speed piecewise linear function is as follows: The target grate cycle and pusher speed are collected over historical periods. Based on the distribution of the target grate cycle and pusher speed, a suitable function form is selected for fitting. For example, linear functions, polynomial functions, exponential functions, etc., can be selected to obtain the speed piecewise linear function.

[0069] In this optional implementation, obtaining the first target grate cycle based on the feeder speed setpoint and the speed piecewise linear function includes: plotting the speed piecewise linear function on a two-dimensional coordinate system, where the horizontal axis and vertical axis of the two-dimensional coordinate system represent the target grate cycle and the feeder speed, respectively; finding the target grate cycle on the horizontal axis; and finding the target grate cycle on the vertical axis that corresponds to both the speed piecewise linear function and the feeder speed setpoint, thereby obtaining the first target grate cycle.

[0070] Optionally, the above-mentioned method of obtaining the first target grate cycle based on the feeder speed setpoint and the speed piecewise linear function includes: obtaining the piecewise linear function expression based on the speed piecewise linear function; and substituting the feeder speed setpoint into the piecewise linear function expression to obtain the first target grate cycle.

[0071] The optional implementation provides a reliable method for obtaining the first target grate period by calculating the first target grate period through a velocity piecewise linear function.

[0072] In some optional implementations of this disclosure, the extraction of bed thickness feature values ​​based on the differential pressure control variable sequence values ​​and the furnace bed differential pressure values ​​includes: clustering the differential pressure control variable sequence values ​​using a clustering algorithm to obtain clustered data; extracting sub-variable values ​​belonging to the appropriate bed thickness from the clustered data; inputting the sub-variable values ​​into a pre-trained long short-term memory neural network to obtain the predicted bed differential pressure output by the long short-term memory neural network; and subtracting the predicted bed differential pressure from the actual bed differential pressure to obtain the bed thickness feature values.

[0073] In this optional implementation, the automatic combustion control system needs to control the fuel thickness within the furnace within a suitable range: Excessive fuel thickness prevents primary air from penetrating the fuel layer, affecting fuel drying and making ignition difficult; insufficient fuel thickness leads to insufficient fuel in the furnace, causing rapid combustion and making it difficult to maintain a suitable furnace temperature. Since there are no sensors directly measuring the fuel layer thickness within the furnace, only the furnace pressure difference can be measured. There are pressure measuring points at the top of the fuel layer and at the lower air chamber. The fuel layer thickness is indirectly reflected by calculating the pressure difference between the upper and lower sides of the furnace.

[0074] In this optional implementation, the pressure difference of the material bed is affected by two key factors: the thickness of the material bed and the real-time flow rate of the primary air. The larger the real-time flow rate of the primary air, the larger the pressure difference of the furnace material bed. The thicker the furnace material bed, the larger the pressure difference of the material bed. Therefore, the thickness of the material bed cannot be directly reflected by the pressure difference of the furnace material bed. A material bed thickness feature extraction model is needed to eliminate the influence of the real-time flow rate of the primary air from the pressure difference of the material bed, and the extracted material bed thickness feature is used as the basis to adjust the pusher speed and the cycle of each grate section.

[0075] In this optional implementation, the above-mentioned clustering algorithm is used to cluster the control variable sequence values ​​to obtain clustered data including: using the K-means clustering algorithm, the control variable sequence values ​​are divided into three categories. Specifically, by drawing a scatter plot representing the relationship between the primary blower frequency and the furnace material bed pressure difference value, they are classified into three types: material bed too thick, material bed too thin, and material bed moderate.

[0076] In this optional implementation, the extraction of sub-variable values ​​belonging to the moderate material layer type from the clustered data includes: selecting data with moderate material layer, performing data preprocessing (data cleaning, filling missing values), and then obtaining the sub-variable values.

[0077] In this optional implementation, the pre-trained long short-term memory neural network is a model used to predict the pressure difference value of the furnace bed (i.e., the predicted pressure difference value of the bed). The bed thickness feature value can be obtained by predicting the pressure difference value of the bed. Specifically, the input variables for the Long Short-Term Memory Neural Network are: primary air frequency, secondary air frequency, furnace negative pressure left, furnace negative pressure right, drying grate I primary air damper A opening, drying grate I primary air damper B opening, drying grate I primary air damper C opening, combustion grate I primary air damper A opening, combustion grate I primary air damper B opening, combustion grate I primary air damper C opening, combustion grate II primary air damper A opening, combustion grate II primary air damper B opening, combustion grate II primary air damper C opening, combustion grate III primary air damper A opening, combustion grate III primary air damper B opening, combustion grate III primary air damper C opening, burnout grate I primary air damper A opening, burnout grate I primary air damper B opening, burnout grate I primary air damper C opening, burnout grate II primary air damper A opening, burnout grate II primary air damper B opening, burnout grate II primary air damper C opening. The output variable of the long short-term memory neural network is the predicted pressure difference of the material layer. From the perspective of process principle, the output of the long short-term memory neural network is regarded as the predicted pressure difference of the material layer corresponding to the current air volume when the material layer thickness is moderate.

[0078] In this optional implementation, the furnace bed pressure difference value is the actual bed pressure difference value, and the bed thickness characteristic = furnace bed pressure difference value - long short-term memory neural network output variable (the furnace bed pressure difference value corresponding to the current air volume when the bed thickness is moderate). When the actual bed pressure difference is inconsistent with the bed pressure difference output by the long short-term memory neural network under the condition of "moderate bed thickness", it can be considered that the bed thickness is too thick or too thin, and the difference between the two reflects the degree of bed thickness.

[0079] The optional implementation provides a method for obtaining the material layer thickness feature value. By using a long short-term memory neural network to obtain the predicted material layer pressure difference value, the material layer thickness feature value is obtained, which improves the reliability and accuracy of obtaining the material layer thickness feature value.

[0080] In some optional implementations of this disclosure, the above-mentioned calculation of the second target grate cycle based on the material layer thickness characteristic value includes: obtaining a material layer piecewise linear function that characterizes the correspondence between the material layer thickness characteristic value and the target grate cycle; and obtaining the second target grate cycle based on the material layer thickness characteristic value and the material layer piecewise linear function.

[0081] In this optional implementation, the material layer piecewise linear function is a piecewise linear function establishing a linear relationship between input and output variables using big data technology. The process of obtaining the material layer piecewise linear function is as follows: Collect the characteristic values ​​of the material layer thickness and the target grate cycle from historical time periods. Based on the distribution of the material layer thickness characteristic values ​​and the target grate cycle, select a suitable function form for fitting. For example, linear functions, polynomial functions, exponential functions, etc., can be selected to obtain the material layer piecewise linear function. The calculation method for the material layer thickness characteristic has been explained in the above embodiments and will not be repeated here.

[0082] In this optional implementation, obtaining the second target grate cycle based on the material layer thickness feature value and the material layer polygonal function includes: plotting the material layer polygonal function on a two-dimensional coordinate system, wherein the horizontal axis and vertical axis of the two-dimensional coordinate system represent the material layer thickness feature value and the target grate cycle, respectively; finding the material layer thickness feature value on the horizontal axis; and finding the target grate cycle that corresponds to both the material layer polygonal function and the material layer thickness feature value on the vertical axis, thereby obtaining the second target grate cycle.

[0083] Optionally, the above-mentioned method of obtaining the second target grate cycle based on the material layer thickness characteristic value and the material layer piecewise linear function includes: obtaining the piecewise linear function expression based on the material layer piecewise linear function; and substituting the material layer thickness characteristic value into the piecewise linear function expression to obtain the second target grate cycle.

[0084] The optional implementation provides a method for obtaining the second target grate period by calculating the second target grate period through a broken line function of the material layer, thus providing a reliable implementation method for obtaining the second target grate period.

[0085] In some optional implementations of this disclosure, the above-mentioned calculation of the target grate cycle given value based on the flow prediction value, the first target grate cycle, the second target grate cycle, and the third target grate cycle includes: obtaining a flow piecewise linear function characterizing the correspondence between the main steam flow value and the target grate cycle; obtaining the fourth target grate cycle based on the flow prediction value and the flow piecewise linear function; and performing a weighted summation of the first target grate cycle, the second target grate cycle, the third target grate cycle, and the fourth target grate cycle to obtain the target grate cycle given value.

[0086] In this optional implementation, the flow rate piecewise linear function is a piecewise linear function establishing a linear relationship between input and output variables using big data technology. The process of obtaining the flow rate piecewise linear function is as follows: Collect the main steam flow rate and target grate cycle from historical time periods. Set the main steam flow rate and target grate cycle as a correspondence with a target delay. Under this correspondence, calibrate the points corresponding to the main steam flow rate and target grate cycle (each point represents a total air volume flow rate and a fuel pusher speed). Based on the distribution of all points, select a suitable function form for fitting. For example, linear functions, polynomial functions, exponential functions, etc., can be selected to obtain the flow rate piecewise linear function. The method for obtaining the target delay is the same as described in the above embodiments, only the dataset (in this example, the dataset is a collection of multiple total air volume values ​​and fuel pusher speeds) is different, and will not be elaborated further here. Figure 2 As shown, the fourth target grate cycle (not shown in the figure) can be obtained through the predicted flow rate and the flow rate piecewise linear function F4(x). Figure 2 After weighting the first target grate cycle, the second target grate cycle, the third target grate cycle, and the fourth target grate cycle, the given value of the target grate cycle is obtained. Figure 2 (not shown in the figure), the target grate cycle setpoint is entered into PID control.

[0087] In this optional implementation, obtaining the fourth target grate cycle based on the predicted flow rate and the flow rate piecewise linear function includes: plotting the flow rate piecewise linear function on a two-dimensional coordinate system, where the horizontal and vertical axes of the two-dimensional coordinate system represent the main steam flow rate and the target grate cycle, respectively; finding the predicted flow rate on the horizontal axis; and finding the target grate cycle that corresponds to both the flow rate piecewise linear function and the predicted flow rate on the vertical axis, thereby obtaining the fourth target grate cycle.

[0088] Optionally, the above-mentioned method of obtaining the fourth target grate cycle based on the flow prediction value and the flow piecewise linear function includes: obtaining the piecewise linear function expression based on the flow piecewise linear function; and substituting the flow prediction value into the piecewise linear function expression to obtain the fourth target grate cycle.

[0089] In this embodiment, the weights of the first target grate cycle, the second target grate cycle, the third target grate cycle, and the fourth target grate cycle in the weighted summation can be set as needed.

[0090] The optional implementation provides a method for obtaining the target grate cycle setpoint by calculating the fourth target grate cycle through a flow rate piecewise linear function, and obtaining the target grate cycle setpoint by weighted summing of the first, second, third, and fourth target grate cycles, thus providing a reliable implementation method for obtaining the target grate cycle setpoint.

[0091] In some optional implementations of this disclosure, the above-mentioned dynamic correction of the target grate cycle setpoint based on the flow deviation to obtain and use the target grate cycle to control the target grate includes: inputting the flow deviation and the target grate cycle setpoint to a proportional-integral-derivative controller to obtain the target grate cycle output by the proportional-integral-derivative controller; and using the target cycle to control the target grate.

[0092] In this optional implementation, the proportional-integral-derivative controller adopts a PID closed-loop control architecture, such as... Figure 2 or Figure 3 As shown, the flow deviation between the main steam flow setpoint and the flow prediction is used as the control input. The target grate cycle setpoint is dynamically corrected through the adjustment mechanism of the proportional-integral-derivative controller to obtain the target grate cycle.

[0093] In this optional implementation, the proportional-integral-derivative controller control principle is as shown in equation (5), and the target period Y(s) of the target grate is obtained.

[0094] (5) In equation (5), FF(s) is the periodic reference value, which is also the target grate period setpoint, E(s) is the flow deviation, and K... p Ti is the proportionality coefficient, and T is the integral coefficient. d These are the differential coefficients.

[0095] The optional implementation provides a method for obtaining the target grate's target cycle by inputting the flow deviation and the target grate cycle setpoint into a proportional-integral-derivative controller to obtain the target grate's target cycle output by the proportional-integral-derivative controller. This method is simple and convenient to implement.

[0096] Optionally, before controlling the speed of the fuel pusher using the target grate's target cycle, the above-mentioned fuel pusher control method further includes limiting the speed value output by the proportional-integral-derivative controller before obtaining the target grate's target cycle, thereby obtaining the target grate's target cycle. Specifically, as shown in... Figure 2 As shown, it should be noted that the final output values ​​of all control loops (pusher speed, primary fan frequency, secondary fan frequency, and grate cycle of each section) need to be limited and speed-limited as a safety protection measure after final output. This is to prevent the final output value from exceeding the normal range due to sensor jump failure or other abnormal conditions, which could cause a production accident. The specific limit value should be adjusted by the operator according to the normal range value.

[0097] Further reference Figure 4As an implementation of the methods shown in the above figures, this disclosure provides an embodiment of a grate cycle control device, which is similar to... Figure 1 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0098] like Figure 4 As shown, the grate cycle control device 400 provided in this embodiment includes: an acquisition unit 401, a prediction calculation unit 402, a deviation calculation unit 403, a first cycle calculation unit 404, an extraction unit 405, a second cycle calculation unit 406, a third cycle calculation unit 407, a setpoint calculation unit 40408, and a control unit 40409. The acquisition unit 401 can be configured to acquire, during a historical period, the parameter sequence values ​​of control parameters related to the main steam flow rate in the automatic combustion control system, the current flow rate value of the main steam flow rate, the flow rate setpoint value of the main steam flow rate, the pusher speed setpoint value, the furnace bed pressure difference value, and, during a historical period, the pressure difference control variable sequence values ​​related to the furnace bed pressure difference height, and the calculated bed thickness value related to the target grate bed thickness. The prediction calculation unit 402 can be configured to calculate the predicted flow rate value of the main steam flow rate based on the parameter sequence values ​​and the current flow rate value. The deviation calculation unit 403 can be configured to calculate the flow deviation based on the flow rate setpoint value and the predicted flow rate value. The first cycle calculation unit 404 can be configured to calculate the first target grate cycle based on the feeder speed setpoint. The extraction unit 405 can be configured to extract the material layer thickness feature value based on the differential pressure control variable sequence value and the furnace material layer differential pressure value. The second cycle calculation unit 406 can be configured to calculate the second target grate cycle based on the material layer thickness feature value. The third cycle calculation unit 407 can be configured to calculate the third target grate cycle based on the calculated material layer thickness value. The setpoint calculation unit 40408 can be configured to calculate the target grate cycle setpoint based on the flow prediction value, the first target grate cycle, the second target grate cycle, and the third target grate cycle. The control unit 40409 can be configured to dynamically correct the target grate cycle setpoint based on the flow deviation, obtain and use the target grate cycle to control the target grate.

[0099] In this embodiment, the specific processing and technical effects of the following components in the grate cycle control device 400—namely, the acquisition unit 401, the prediction calculation unit 402, the deviation calculation unit 403, the first cycle calculation unit 404, the extraction unit 405, the second cycle calculation unit 406, the third cycle calculation unit 407, the setpoint calculation unit 408, and the control unit 409—can be found in reference [reference needed]. Figure 1The relevant descriptions of steps 101, 102, 103, 104, 105, 106, 107, 108, and 109 in the corresponding embodiments will not be repeated here.

[0100] In some embodiments of this disclosure, the target grate is a combustion grate, and the calculated material layer thickness includes a combustion grate thickness value and a burnout grate thickness value. The third cycle calculation unit 407 is configured to: calculate a first sub-grate cycle based on the combustion grate thickness value; calculate a second sub-grate cycle based on the burnout grate thickness value; and obtain a third target grate cycle based on the first sub-grate cycle and the second sub-grate cycle.

[0101] In some embodiments of this disclosure, the first cycle calculation unit 404 is configured to: obtain a velocity piecewise linear function characterizing the relationship between the target grate cycle and the pusher speed; and obtain the first target grate cycle based on the pusher speed setpoint and the velocity piecewise linear function.

[0102] In some embodiments of this disclosure, the extraction unit 405 is configured to: cluster the differential pressure control variable sequence values ​​using a clustering algorithm to obtain clustered data; extract sub-variable values ​​belonging to the appropriate material layer from the clustered data; input the sub-variable values ​​into a pre-trained long short-term memory neural network to obtain the predicted material layer differential pressure output by the long short-term memory neural network; and subtract the predicted material layer differential pressure from the actual material layer differential pressure to obtain the material layer thickness feature value.

[0103] In some embodiments of this disclosure, the second cycle calculation unit 406 is configured to: obtain a broken line function representing the relationship between the material layer thickness characteristics and the target grate cycle; and obtain the second target grate cycle based on the material layer thickness characteristic value and the broken line function.

[0104] In some embodiments of this disclosure, the given value calculation unit 408 is configured to: obtain a flow rate piecewise linear function representing the correspondence between the main steam flow rate value and the target grate cycle; obtain a fourth target grate cycle based on the flow rate prediction value and the flow rate piecewise linear function; and perform a weighted summation of the first target grate cycle, the second target grate cycle, the third target grate cycle, and the fourth target grate cycle to obtain a given value for the target grate cycle.

[0105] In some embodiments of this disclosure, the control unit 409 is configured to: input the flow deviation and the target grate cycle setpoint to the proportional-integral-derivative controller to obtain the target grate cycle output by the proportional-integral-derivative controller; and control the target grate using the target cycle.

[0106] The grate cycle control device provided in the embodiments of this disclosure includes an acquisition unit 401 that acquires the parameter sequence values ​​of control parameters related to the main steam flow rate during a historical period of the automatic combustion control system, the current flow rate value of the main steam flow rate, the flow rate setpoint value of the main steam flow rate, the pusher speed setpoint value, the furnace bed pressure difference value, and the pressure difference control variable sequence values ​​related to the furnace bed pressure difference height during a historical period, as well as the calculated bed thickness value related to the target grate bed thickness. A prediction calculation unit 402 calculates the predicted flow rate value of the main steam flow rate based on the parameter sequence values ​​and the current flow rate value. A deviation calculation unit 403 calculates the flow deviation based on the flow rate setpoint value and the predicted flow rate value. A first cycle calculation unit... Unit 404 calculates the first target grate cycle based on the feeder speed setpoint; extraction unit 405 extracts the material layer thickness feature value based on the differential pressure control variable sequence value and the furnace material layer differential pressure value; second cycle calculation unit 406 calculates the second target grate cycle based on the material layer thickness feature value; third cycle calculation unit 407 calculates the third target grate cycle based on the material layer thickness calculation value; setpoint calculation unit 408 calculates the target grate cycle setpoint based on the flow prediction value, the first target grate cycle, the second target grate cycle, and the third target grate cycle; control unit 409 dynamically corrects the target grate cycle setpoint based on the flow deviation, obtains and uses the target grate cycle to control the target grate.

[0107] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0108] Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their patterns are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0109] like Figure 5As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. The RAM 503 may also store various programs and data required for the operation of the device 500. The computing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0110] Multiple components in electronic device 500 are connected to I / O interface 505, including: input unit 506, such as keyboard, mouse, etc.; output unit 507, such as various types of monitors, speakers, etc.; storage unit 508, such as disk, optical disk, etc.; and communication unit 509, such as network card, modem, wireless transceiver, etc. Communication unit 509 allows device 500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0111] The computing unit 501 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as the grate cycle control method. For example, in some embodiments, the grate cycle control method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the grate cycle control method described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to perform the grate cycle control method by any other suitable means (e.g., by means of firmware).

[0112] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0113] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable grate cycle control device, such that when executed by the processor or controller, the program code causes the patterns / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0114] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0115] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0116] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0117] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0118] The foregoing description of specific exemplary embodiments of this disclosure is for illustrative and explanatory purposes. These descriptions are not intended to limit this disclosure to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of this disclosure and their practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of this disclosure, as well as various different choices and variations. The scope of this disclosure is intended to be defined by the claims and their equivalents.

Claims

1. A grate cycle control method, applied to an automatic combustion control system, characterized in that, The method includes: The automatic combustion control system acquires the parameter sequence values ​​of the control parameters related to the main steam flow rate during historical periods, the current flow rate value of the main steam flow rate, the flow rate setpoint value of the main steam flow rate, the pusher speed setpoint value, the furnace bed pressure difference value, the pressure difference control variable sequence values ​​related to the furnace bed pressure difference height during historical periods, and the calculated bed thickness value related to the target grate bed thickness. Based on the parameter sequence values ​​and the current flow rate value, calculate the predicted flow rate of the main steam flow rate; Calculate the flow deviation based on the flow setpoint and the flow prediction. The first target grate cycle is calculated based on the given value of the pusher speed; Based on the differential pressure control variable sequence value and the furnace material layer differential pressure value, the material layer thickness feature value is extracted; Based on the material layer thickness characteristic value, the second target grate cycle is calculated; Based on the calculated material layer thickness, the third target grate cycle is calculated; Based on the predicted flow rate, the first target grate cycle, the second target grate cycle, and the third target grate cycle, calculate the target grate cycle setpoint. Based on the flow deviation, the target grate cycle setpoint is dynamically corrected to obtain and use the target grate cycle to control the target grate.

2. The method according to claim 1, characterized in that, The target grate is a combustion grate, and the calculated material layer thickness includes: the combustion grate thickness and the burnout grate thickness. The calculation of the third target grate cycle based on the calculated material layer thickness includes: calculating the first sub-grate cycle based on the combustion grate thickness; calculating the second sub-grate cycle based on the burnout grate thickness; and obtaining the third target grate cycle based on the first sub-grate cycle and the second sub-grate cycle.

3. The method according to claim 1 or 2, characterized in that, The calculation of the first target grate cycle based on the given value of the pusher speed includes: Obtain a velocity piecewise linear function that represents the relationship between the target grate cycle and the pusher speed; Based on the feeder speed setpoint and the speed piecewise linear function, the first target grate cycle is obtained.

4. The method according to claim 1, characterized in that, The extraction of material layer thickness feature values ​​based on the differential pressure control variable sequence values ​​and the furnace material layer differential pressure values ​​includes: Clustering algorithms are used to cluster the differential pressure control variable sequence values ​​to obtain clustered data. Extract sub-variable values ​​belonging to the appropriate material layer from the clustered data; The sub-variable values ​​are input into a pre-trained long short-term memory neural network to obtain the predicted material layer pressure difference output by the long short-term memory neural network; Subtracting the predicted material layer pressure difference from the actual material layer pressure difference yields the characteristic value of the material layer thickness.

5. The method according to claim 1, characterized in that, The calculation of the second target grate cycle based on the material layer thickness characteristic value includes: Obtain the piecewise linear function representing the relationship between the thickness characteristics of the bed and the target grate cycle; The second target grate cycle is obtained based on the material layer thickness characteristic value and the material layer polygonal function.

6. The method according to claim 1, characterized in that, The calculation of the target grate cycle given value based on the predicted flow rate, the first target grate cycle, the second target grate cycle, and the third target grate cycle includes: Obtain a flow piecewise linear function representing the relationship between the main steam flow rate and the target grate cycle; Based on the predicted flow rate and the flow rate piecewise linear function, the fourth target grate cycle is obtained; The first target grate cycle, the second target grate cycle, the third target grate cycle, and the fourth target grate cycle are weighted and summed to obtain the target grate cycle given value.

7. The method according to claim 1, characterized in that, The step of dynamically correcting the target grate cycle setpoint based on the flow deviation, and obtaining and using the target grate cycle to control the target grate includes: The flow deviation and the target grate cycle setpoint are input into the proportional-integral-derivative controller to obtain the target grate cycle output by the proportional-integral-derivative controller; The target grate is controlled using the target cycle.

8. A grate cycle control device, the device comprising: The acquisition unit is configured to acquire the parameter sequence values ​​of the control parameters related to the main steam flow rate during the historical period of the automatic combustion control system parameter sequence values, the current flow rate value of the main steam flow rate, the flow rate setpoint value of the main steam flow rate, the pusher speed setpoint value, the furnace bed pressure difference value, the pressure difference control variable sequence value related to the furnace bed pressure difference height during the historical period, and the calculated bed thickness value related to the target grate bed thickness. The prediction calculation unit is configured to calculate the predicted flow rate of the main steam flow rate based on the parameter sequence value and the current flow rate value; The deviation calculation unit is configured to calculate the flow deviation based on the flow setpoint and the flow prediction value; The first cycle calculation unit is configured to calculate the first target grate cycle based on the given value of the pusher speed; The extraction unit is configured to extract the material layer thickness feature value based on the differential pressure control variable sequence value and the furnace material layer differential pressure value; The second cycle calculation unit is configured to calculate the second target grate cycle based on the material layer thickness characteristic value; The third cycle calculation unit is configured to calculate the third target grate cycle based on the calculated value of the material layer thickness. The given value calculation unit is configured to calculate a target grate cycle given value based on the predicted flow rate, the first target grate cycle, the second target grate cycle, and the third target grate cycle. The control unit is configured to dynamically correct the target grate cycle setpoint based on the flow deviation, and to obtain and control the target grate using the target grate cycle.

9. An electronic device, characterized in that, include: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

Images