Adaptive optimization start-stop control system for coal pulverizing system of thermal power unit based on aps

Abstract

The present application belongs to the technical field of industrial automation control, and discloses a self-adaptive optimization start-stop control system for a coal pulverizing system of a thermal power unit based on APS, which comprises an APS instruction analysis module, a working condition sensing module, and a section division module.The APS instruction analysis module is used for analyzing APS start-stop instructions, extracting process constraint parameters of the coal pulverizing system, combining with the operating load of the boiler, and generating start-stop control targets and load adjustment boundaries.The working condition sensing module extracts working condition interaction factors reflecting coal types, equipment, and air-pulverized coal interaction by collecting operating condition data, constructs a state-energy-load three-dimensional mapping relationship, and forms a physical side state vector.The section division module constructs a control subspace of the energy flow variation interval of the coal pulverizing system based on the physical side state vector, divides the control subspace into different energy operation sections, calculates the energy release rate of a preset target section and a current section, and generates a start-stop strategy switching signal.The energy utilization efficiency is improved, and the equipment loss is reduced.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology, and more specifically, to an adaptive optimization start-stop control system for pulverizing systems of thermal power units based on APS. Background Technology

[0002] Patent publication number CN112594668A discloses a solution to overheating and overpressure issues during the start-up and shutdown of a coal mill in a thermal power unit. Based on the accumulated coal dust in the pulverizer during startup, the midpoint temperature rise of the supercritical once-through boiler is determined. Based on the correlation between the determined midpoint temperature rise and the feedwater flow rate, a target feedwater flow rate increase corresponding to the midpoint temperature rise during mill startup and shutdown is determined. This is achieved by sending a feedwater command corresponding to the target feedwater flow rate to the thermal power unit, allowing the unit to increase the target feedwater flow rate in advance. This invention pre-increases the water flow rate based on the accumulated coal dust in the pulverizer, preventing excessive deviation in the coal-water ratio during mill startup. This quickly suppresses midpoint temperature changes, reduces the impact on superheated temperatures, and allows the thermal power unit to stabilize rapidly. It also improves the unit's automation control level, ensuring safe and stable operation and guaranteeing the unit's load response speed.

[0003] The existing adaptive optimization start-stop control system for pulverizing power units has the following main problems:

[0004] In existing technologies, only the total energy consumption of the system is monitored and adjusted, resulting in insufficiently refined control strategies that are difficult to adapt to complex operating conditions. Under complex and changing conditions, the lack of flexible control mechanisms in existing technologies leads to low operating efficiency of the pulverizing system, and may even cause problems such as overload or inefficient operation. In the current start-stop control of thermal power unit pulverizing systems, start-stop commands are typically responded to based on fixed threshold triggering mechanisms or manually set timing strategies, ignoring the impact of factors such as coal type changes, equipment condition deterioration, and air-coal coupling disturbances on the energy release path and response rhythm during actual operation. Once a start-stop command is triggered, it is forcibly executed without dynamic adjustment of the response time, easily causing energy consumption fluctuations, heat load shocks, and even inducing boiler instability. The system lacks quantitative analysis of the energy release behavior of the current operating segment and the target segment, resulting in blind and lagging control behavior; the system is easily affected by disturbance amplification effects when executing start-stop operations, leading to increased execution deviations.

[0005] In view of this, the present invention proposes an adaptive optimization start-stop control system for the pulverizing system of thermal power units based on APS to solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: an adaptive optimization start-stop control system for a thermal power unit pulverizing system based on APS, comprising:

[0007] The APS instruction parsing module is used to parse APS start-stop instructions, extract process constraint parameters of the pulverizing system, and generate start-stop control targets and load adjustment boundaries in combination with boiler operating load.

[0008] The operating condition sensing module collects operating condition data, extracts operating condition interaction factors that reflect the interaction between coal type, equipment and air-coal mixture, constructs a three-dimensional mapping relationship between state, energy and load, and forms a physical-side state vector.

[0009] The segment division module constructs a control subspace for the energy rheology range of the pulverizing system based on the physical side state vector, divides it into different energy operation segments, calculates the energy release rate of the preset target segment and the current segment, and generates start-stop strategy switching signals.

[0010] The start-stop correction module combines the dynamic change curve of the APS start-stop command with the current device health status data to construct a real-time optimization space for the start-stop strategy parameters and perform self-iterative correction of the start-stop strategy parameters.

[0011] The performance evaluation module receives start-stop strategy switching signals, constructs an actuator coupling behavior matrix, identifies transient coupling paths between multi-source controls, introduces a disturbance suppression mechanism, absorbs coal quality disturbances and wind pressure fluctuations in real time, and evaluates the effectiveness of the start-stop execution process.

[0012] Preferably, the method for obtaining the process constraint parameters includes:

[0013] The system receives structured start-stop commands issued by the APS, performs semantic recognition and parsing on the start-stop commands, and extracts process constraint parameters. The process constraint parameters include minimum air-to-coal ratio, maximum number of start-stop cycles per hour, priority start-stop grouping of coal mills, start-stop response delay time, upper limit of coal mill shell temperature, and lower limit of main steam pressure.

[0014] Preferably, the method for obtaining the start / stop control target and load adjustment boundary includes:

[0015] The APS start-stop command types are analyzed, and the process constraint parameters of the pulverizing system are extracted from them. Combined with the current operating load data of the boiler, a weighted multi-objective optimization function is constructed, and the start-stop control objective is finally output.

[0016] The load adjustment boundary is derived based on the current operating load state of the boiler and is used to constrain the allowable operating parameter range of the pulverizing system during start-up and shutdown: the initial physical limit load boundary of the equipment is determined according to the design parameters of the core equipment of the pulverizing system; the initial physical limit load boundary is narrowed and corrected based on the safety operation indicators in the process constraint parameters; the corrected initial physical limit load boundary is then verified a second time by combining the minimum stable combustion feed rate and the maximum allowable feed rate corresponding to the current operating load of the boiler, and the final load adjustment boundary is generated.

[0017] Preferably, the method for obtaining the working condition interaction factor includes:

[0018] Operating condition data is collected, including online coal quality analysis data, coal mill operating status data, and air pressure and velocity data. The data is preprocessed through standardization and anomaly removal. Based on principal component analysis, operating condition interaction factors reflecting coal type characteristics, equipment response behavior, and the coupling relationship between air and coal conveying are extracted. These operating condition interaction factors include coal type crushing efficiency factor, air-to-coal ratio deviation factor, and equipment dynamic response factor, which characterize the coupling characteristics of the pulverizing system under different operating conditions.

[0019] Preferably, the method for obtaining the physical-side state vector includes:

[0020] Based on the extracted operating condition interaction factors, coal quality rheological parameters, equipment health status parameters, and gas-solid transport status parameters are further integrated to construct a multi-dimensional state vector describing the operating characteristics of the pulverizing system; and based on this multi-dimensional state vector, state-energy mapping model and state-load mapping model are established respectively.

[0021] The energy consumption level required per unit load is calculated using the state-energy mapping model, and the maximum load regulation capacity of the system under the current operating conditions is calculated using the state-load mapping model. The output forms the physical-side state vector.

[0022] Preferably, the method for obtaining the energy operating range includes:

[0023] Based on the equipment operating status corresponding to each energy operating segment, the energy release rate of the energy operating segment per unit time is calculated; using the physical side state vector and equipment operating status, a control subspace of the energy rheological range of the pulverizing system is constructed, which reflects the energy change law and boundary of the system under different operating conditions; based on the numerical range of the energy release rate, the control subspace of the energy rheological range of the pulverizing system is divided into different energy operating segments.

[0024] Preferably, the method for generating the start / stop strategy switching signal includes:

[0025] For the target operating segment, when the start / stop command is issued, the system sets the target energy release rate to guide the direction of control execution; it presets the energy release rate of the current operating segment per unit time and calculates the difference in energy release rate between the current segment and the target segment during operation;

[0026] A dynamic delay time function is introduced to calculate the execution delay time of the current start-stop strategy and control the execution rhythm of the start-stop action; a preset dynamic delay time threshold is set. When the execution delay time of the current start-stop strategy is greater than or equal to the preset dynamic delay time threshold, it is determined that the execution conditions are met and a start-stop strategy switching signal is generated.

[0027] Preferably, the method for constructing the real-time optimization space includes:

[0028] Collect the historical start-stop command sequence issued by APS, extract the feature information of the historical APS start-stop commands, and form the dynamic change curve of APS start-stop commands; obtain the current equipment health status data and construct the equipment health status vector;

[0029] Based on the dynamic change curve of APS start / stop commands and the device health status vector, a real-time optimization space for start / stop strategy parameters is constructed. The optimal combination of start / stop strategy parameters is selected in the real-time optimization space using a greedy algorithm, and the start / stop strategy parameters of the current start / stop strategy are iteratively corrected.

[0030] Preferably, the method for identifying the transient coupling path includes:

[0031] The control input signals and response output data of each actuator in the pulverizing system are acquired. Based on the response relationship between the input and output of each actuator, an actuator coupling behavior matrix is ​​constructed. The actuator coupling behavior matrix is ​​normalized, and the main coupling path is extracted by the eigenvalue decomposition method. Each actuator is modeled as a node, and the coupling relationship is modeled as an edge. A directed graph of coupling relationship is constructed, and transient coupling paths existing during start-stop switching are identified.

[0032] Preferably, the method for evaluating the effectiveness of the start-stop execution process includes:

[0033] A disturbance suppression mechanism is introduced to identify and suppress the instability caused by coal quality disturbances and wind pressure fluctuations to the system in real time during the start-up and shutdown strategy execution, and to evaluate the effect of start-up and shutdown execution. The disturbance suppression mechanism includes identifying the abrupt behavior characteristics of coal particle size, moisture, and volatile matter parameters and the dynamic fluctuation characteristics of the air supply system by collecting online coal quality analysis data, wind pressure and wind speed sensor data and coal mill load data, and establishing a disturbance source feature vector.

[0034] A system disturbance propagation model including control input and disturbance input is constructed to extract the influence path of disturbance on key output variables of the pulverizing system; based on the disturbance source feature vector, a disturbance suppression strategy is implemented.

[0035] By constructing an efficiency scoring function, the efficiency indicators of the pulverizing system are quantified, thereby enabling the evaluation of the start-up and shutdown processes.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] This invention constructs a physical-side state vector and combines it with equipment operating states such as mill power, pulverized coal output rate, and air-to-pulverized coal ratio to accurately calculate the energy release rate of each energy operating segment. By constructing a control subspace for the energy rheology interval, it flexibly divides the system into multiple energy operating segments based on the real-time energy release rate and different operating states. Each energy operating segment has a clearly defined energy release rate boundary and control strategy, ensuring that the system can accurately match energy demand at different stages to achieve optimal operating conditions. Through real-time monitoring of the energy release rate and equipment status, the control objectives of each segment can be flexibly adjusted according to actual operating conditions, thereby achieving optimal energy allocation and utilization.

[0038] By calculating the difference between the current energy release rate and the target rate in real time and introducing a disturbance intensity factor for normalization, the execution delay time of the start-stop control strategy is dynamically adjusted. This mechanism significantly improves the control system's adaptability to fluctuations in operating conditions, making start-stop operations more flexible and controllable, and effectively avoiding control shocks and system oscillations caused by excessively rapid responses. By setting a dynamic delay time threshold, the start-stop strategy switching signal is generated only when the current energy behavior meets the strategy execution conditions, effectively preventing invalid start-stops or unnecessary adjustments, and improving the rationality of control strategy execution and system stability. It can absorb disturbances caused by factors such as coal quality changes and wind pressure fluctuations in real time, achieving active suppression and response optimization of typical disturbance sources, and enhancing the system's robustness to complex operating conditions. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the adaptive optimization start-stop control system structure of the APS-based pulverizing system for thermal power units according to the present invention.

[0040] Figure 2 This is a schematic diagram of the adaptive optimization start-stop control method for the pulverizing system of a thermal power unit based on APS according to the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1

[0043] Please see Figure 1 As shown, this embodiment further illustrates the adaptive optimization start-stop control system for the pulverizing system of thermal power units based on APS proposed in this invention, including:

[0044] In traditional pulverizing systems of thermal power units, the start-up and shutdown control typically employs fixed threshold triggering mechanisms or manually set timing strategies. This approach often overlooks the impact of actual operating condition changes on the system, resulting in insufficiently refined control strategies that struggle to cope with complex and dynamic operating environments. Existing control strategies are often unable to adapt to the influence of factors such as coal type changes, equipment aging, and air-coal coupling disturbances on energy release paths and control responses, leading to low system efficiency and even causing overload or inefficient operation.

[0045] Specifically, existing technologies have the following shortcomings: Inaccurate energy consumption monitoring: In traditional systems, only total energy consumption is monitored and regulated, without considering the energy release behavior of individual subsystems and different energy operating segments. This coarse energy regulation method is difficult to effectively adapt to different operating conditions, cannot accurately control energy flow, leading to energy waste and reduced system efficiency.

[0046] Lack of flexible control mechanisms: Under complex and changing operating conditions, traditional start-stop control methods are often based on fixed threshold triggering mechanisms, lacking flexible adjustment space. When factors such as changes in coal quality, fluctuations in air-to-coal ratio, and changes in equipment status occur, the system cannot dynamically adjust the start-stop response time, resulting in lagging or overly aggressive control actions. This rigid control method may lead to over-response or delayed response, seriously affecting the efficiency of the pulverizing system and even adversely affecting boiler operation.

[0047] Unable to handle disturbances: Because traditional control strategies do not consider the impact of disturbance sources, such as changes in coal quality and fluctuations in air pressure, start-up and shutdown operations are often amplified by external disturbances, resulting in increased system execution deviations. This control method, lacking a disturbance suppression mechanism, makes the system poorly adaptable to complex operating conditions and may even lead to boiler instability, causing system failures or performance degradation.

[0048] Insufficient quantitative analysis of energy release behavior: Existing start-stop control systems lack quantitative analysis of the energy release behavior of the pulverizing system, especially the division of different energy operating zones and the dynamic calculation of energy release rates. Traditional systems often cannot clearly identify the switching and transition between different energy zones, resulting in start-stop operations without clear objectives and basis, and control behavior exhibiting blindness and lag.

[0049] Ineffective handling of complex operating conditions: Under the influence of factors such as changes in coal type, equipment aging, and air-coal coupling disturbances, existing start-stop control strategies often cannot adjust the energy release rate and start-stop strategy in real time and accurately, leading to inefficient system operation, or even system overload or inefficient operation. Traditional fixed control strategies cannot cope with changes in energy demand under complex operating conditions, resulting in energy waste and system load fluctuations.

[0050] To effectively address the aforementioned problems, this invention proposes an adaptive optimization start-stop control system for the pulverizing system of thermal power units based on APS, comprising:

[0051] The APS instruction parsing module is used to parse APS start-stop instructions, extract process constraint parameters of the pulverizing system, and generate start-stop control targets and load adjustment boundaries in combination with boiler operating load.

[0052] The operating condition sensing module collects operating condition data, extracts operating condition interaction factors that reflect the interaction between coal type, equipment and air-coal mixture, constructs a three-dimensional mapping relationship between state, energy and load, and forms a physical-side state vector.

[0053] The segment division module constructs a control subspace for the energy rheology range of the pulverizing system based on the physical side state vector, divides it into different energy operation segments, calculates the energy release rate of the preset target segment and the current segment, and generates start-stop strategy switching signals.

[0054] The start-stop correction module combines the dynamic change curve of the APS start-stop command with the current device health status data to construct a real-time optimization space for the start-stop strategy parameters and perform self-iterative correction of the start-stop strategy parameters.

[0055] The performance evaluation module receives start-stop strategy switching signals, constructs an actuator coupling behavior matrix, identifies transient coupling paths between multi-source controls, introduces a disturbance suppression mechanism, absorbs coal quality disturbances and wind pressure fluctuations in real time, and evaluates the effectiveness of the start-stop execution process.

[0056] Methods for obtaining process constraint parameters include:

[0057] The system receives structured start-stop commands issued by the APS, performs semantic recognition and parsing on the commands, and extracts process constraint parameters. The start-stop commands include the target start-stop status, the target load range for scheduling, the effective time period, and process constraint parameters related to the pulverizing system. The process constraint parameters include the minimum air-to-coal ratio, the maximum number of start-stop cycles per hour, the priority start-stop grouping of the coal mill, the start-stop response delay time, the upper limit of the coal mill shell temperature, and the lower limit of the main steam pressure.

[0058] Methods for obtaining start / stop control targets and load regulation boundaries include:

[0059] The APS start-up and shutdown command types are analyzed, and process constraint parameters of the pulverizing system (such as upper limit of mill bearing temperature, lubricating oil pressure threshold, pulverized coal fineness requirements, and minimum pulverized coal flow rate required for stable boiler combustion) are extracted. Combined with the current boiler operating load data, a weighted multi-objective optimization function is constructed with start-up time, energy consumption, and load fluctuation (deviation between actual boiler load and preset target load) as optimization objectives. The final output start-up and shutdown control objectives include time objectives (stage completion time, such as reaching rated output within 10 minutes after mill start-up), energy efficiency objectives (electricity consumption threshold per unit pulverized coal output), and stability objectives (pulverized coal flow rate fluctuation rate, such as within ±2%). Boiler operating load data includes boiler evaporation rate, main steam pressure, reheater outlet temperature, and actual coal consumption.

[0060] APS Instruction Analysis: APS (Automatic Power Generation Control) start-stop instructions include the equipment target load and start-stop stage (such as cold start, hot start, normal shutdown, etc.); the current start-stop stage is identified through the instruction code (such as "pulverizing system preheating" "coal mill start-up sequence", etc.).

[0061] The load adjustment boundary is derived based on the current operating load state of the boiler and is used to constrain the allowable operating parameter range of the pulverizing system during start-up and shutdown: the initial physical limit load boundary of the equipment is determined according to the design parameters of the core equipment of the pulverizing system; the initial physical limit load boundary is narrowed and corrected based on the safety operation indicators in the process constraint parameters; the corrected initial physical limit load boundary is then verified a second time by combining the minimum stable combustion feed rate and the maximum allowable feed rate corresponding to the current operating load of the boiler, and the final load adjustment boundary is generated.

[0062] Methods for obtaining operating condition interaction factors include:

[0063] Operating condition data is collected, including online coal quality analysis data, coal mill operating status data, and air pressure and velocity data. The data is preprocessed through standardization and anomaly removal. Based on principal component analysis, operating condition interaction factors reflecting coal type characteristics, equipment response behavior, and the coupling relationship between air and coal conveying are extracted. These operating condition interaction factors include coal type crushing efficiency factor, air-to-coal ratio deviation factor, and equipment dynamic response factor, which characterize the coupling characteristics of the pulverizing system under different operating conditions.

[0064] Methods for obtaining the physical-side state vector include:

[0065] Based on the extracted operating condition interaction factors, coal rheological parameters, equipment health status parameters, and gas-solid transport status parameters are further integrated. The coal rheological parameters include coal powder particle size, moisture, and ash content; the equipment health status parameters include motor temperature rise margin, current fluctuation rate, and start-stop load ratio; and the gas-solid transport status parameters include primary air velocity uniformity, pipeline pressure loss, and air-to-powder ratio deviation. A multidimensional state vector describing the operating characteristics of the pulverizing system is constructed; and a state-energy mapping model and a state-load mapping model are established based on this multidimensional state vector.

[0066] The energy consumption level required per unit load is calculated using the state-energy mapping model, and the maximum load regulation capacity of the system under the current operating conditions is calculated using the state-load mapping model. The output forms the physical-side state vector.

[0067] Methods for obtaining energy operating ranges include:

[0068] The energy operation segment is used to represent the energy release capacity and efficiency state of the pulverizing system under specific operating conditions per unit time. It is the mapping result of the physical side state vector in the control subspace. Due to the dynamic changes of factors such as coal type difference, equipment wear degree, and air-coal matching relationship, the energy fluctuation of the system exhibits nonlinear and piecewise continuous characteristics, and its implicit structure needs to be extracted.

[0069] Based on the equipment operating status (such as pulverizer power, pulverized coal output rate, air-to-pulverized coal ratio, etc.) corresponding to each energy operation section, calculate the energy release rate per unit time for each energy operation section. ;

[0070] The energy release rate is: ;in, Indicates the time point of the coal mill Real-time power consumption; Indicates the time point of the coal mill The grinding efficiency is to effectively grind coal powder into particles that meet the preset process requirements. Indicates the time point of the coal mill Output of pulverized coal;

[0071] By utilizing the physical state vector and equipment operating status, a control subspace for the energy rheology range of the pulverizing system is constructed. This control subspace reflects the energy change pattern and boundaries of the system under different operating conditions. Based on the numerical range of the energy release rate, the control subspace for the energy rheology range of the pulverizing system is divided into different energy operating segments.

[0072] The energy operating section is defined as follows: ;in, Indicates the first One energy operation zone; Indicates the first The lower limit of the energy release rate of an energy operating section, that is, the minimum allowable energy release rate of that section; Indicates the first The upper limit of the energy release rate of an energy operating section, that is, the maximum allowable energy release rate of that section; An index representing the energy operating segment; Indicates the total number of energy operation sections;

[0073] This solution addresses the following problems with existing technologies: Existing technologies often only monitor and regulate the total energy consumption of the system, failing to refine the specific energy release rate for each operating segment. This results in insufficiently precise control strategies, making it difficult to effectively adapt to complex operating conditions. Under complex and changing operating conditions (such as changes in coal quality and load fluctuations), existing technologies lack flexible control mechanisms, leading to low operating efficiency of the pulverizing system and potentially causing problems such as overload or inefficient operation.

[0074] The advantages over existing technologies are as follows: By constructing a physical-side state vector and combining it with equipment operating conditions such as mill power, pulverized coal output rate, and air-to-pulverized coal ratio, the energy release rate of each energy operating segment can be accurately calculated. By constructing a control subspace for the energy rheology interval, multiple energy operating segments are flexibly divided according to the system's real-time energy release rate and different operating states. Each energy operating segment has a clear energy release rate boundary and control strategy, ensuring that the system can accurately match energy demand at different stages and achieve optimal operating conditions. By monitoring the energy release rate and equipment status in real time, the control objectives of each segment can be flexibly adjusted according to actual operating conditions, thereby achieving optimal energy allocation and utilization.

[0075] Methods for generating start / stop policy switching signals include:

[0076] For the target operating segment, the system sets the target energy release rate to [value] when the start / stop command is issued. Guide and control the execution direction; preset the energy release rate of the current operating segment per unit time as During operation, the energy release rate difference between the current segment and the target segment is calculated in real time; the energy release rate difference is: ;

[0077] A dynamic delay time function is introduced to calculate the execution delay time of the current start-stop strategy and control the execution rhythm of start-stop actions;

[0078] The dynamic delay time function is: ;in, Indicates the execution delay time of the current start / stop policy; This indicates the base delay duration of the preset start-stop strategy, which refers to the fixed delay time under no disturbance or baseline conditions, used to ensure a smooth transition of start-stop actions; This represents the perturbation intensity factor, used to adjust the degree of influence of energy deviation on the delay time;

[0079] A preset dynamic delay time threshold is set. When the execution delay time of the current start / stop strategy is greater than or equal to the preset dynamic delay time threshold, it is determined that the execution conditions are met, and a start / stop strategy switching signal is generated.

[0080] The following problems in existing technologies have been solved: In the current start-up and shutdown control of pulverizing systems in thermal power units, start-up and shutdown command responses are usually based on fixed threshold triggering mechanisms or manually set timing strategies. This ignores the impact of factors such as coal type changes, equipment condition deterioration, and air-coal coupling disturbances on the energy release path and response rhythm during actual operation. Once a start-up or shutdown command is triggered, it is forcibly executed without dynamic adjustment of the response time, which can easily cause energy consumption fluctuations, heat load shocks, and even induce boiler instability. The system lacks quantitative analysis of the energy release behavior of the current operating section and the target section, resulting in blind and lagging control behavior. The system is also susceptible to disturbance amplification effects when executing start-up and shutdown operations, leading to increased execution deviations.

[0081] The advantages over existing technologies are as follows: By calculating the difference between the current energy release rate and the target rate in real time and introducing a disturbance intensity factor for normalization, the execution delay time of the start-stop control strategy is dynamically adjusted. This mechanism significantly improves the control system's adaptability to fluctuations in operating conditions, making start-stop operations more flexible and controllable, and effectively avoiding control shocks and system oscillations caused by excessively rapid responses. By setting a dynamic delay time threshold, the start-stop strategy switching signal is generated only when the current energy behavior meets the strategy execution conditions, effectively preventing invalid start-stops or unnecessary adjustments, and improving the rationality of control strategy execution and system stability. It can absorb disturbances caused by factors such as coal quality changes and wind pressure fluctuations in real time, achieving active suppression and response optimization of typical disturbance sources, and enhancing the system's robustness to complex operating conditions.

[0082] Methods for constructing the real-time optimization space include:

[0083] Collect historical start-stop command sequences issued by APS, extract feature information of historical APS start-stop commands, including command target change trend, change slope, duration and target boundary value, to form dynamic change curves of APS start-stop commands; obtain current equipment health status data, including health status data of coal mill, coal feeder, electric actuator and damper equipment in pulverizing system, health status data including current fluctuation, temperature rise, component wear degree and operation stability index, to construct equipment health status vector;

[0084] Based on the dynamic change curve of APS start / stop commands and the equipment health status vector, a real-time optimization space containing start / stop delay time, target air-to-coal ratio, minimum loading threshold strategy parameters is dynamically generated as the adjustable range of control parameters. A greedy algorithm is used to select start / stop strategy parameter combinations in the real-time optimization space and perform self-iterative correction of the start / stop strategy parameters of the current start / stop strategy.

[0085] Methods for identifying transient coupling paths include:

[0086] The control input signals and response output data of each actuator in the pulverizing system are acquired. Based on the response relationship between the input and output of each actuator, an actuator coupling behavior matrix is ​​constructed to reflect the transient behavior coupling characteristics of the actuator. The elements in the actuator coupling behavior matrix are the sensitivity values ​​of each control input to the system output response.

[0087] The actuator coupling behavior matrix is ​​normalized, the main coupling path is extracted using the eigenvalue decomposition method, each actuator is modeled as a node, and the coupling relationship is modeled as an edge, a directed graph of coupling relationship is constructed, and transient coupling paths existing during start-stop switching are identified.

[0088] It should be noted that during the start-up and shutdown of the pulverizing system, multiple actuators (such as the coal mill, coal feeder, and dampers) undergo state changes almost simultaneously. Due to the delay difference between control commands and physical responses, if the linkage between these actuators is not identified and coordinated: the coal feeder may load prematurely before the coal mill starts, easily causing coal blockage; sudden changes in damper opening but an imbalance in the air-to-coal ratio may lead to unstable combustion; rapid adjustments in various control channels may cause problems such as air pressure resonance and power oscillation. Therefore, identifying coupling paths can predict risks before control execution and coordinate the sequence and rate of actions in advance. Different devices have different dynamic response characteristics (e.g., the coal mill has high inertia, and the damper moves quickly). If this difference is ignored when executing start-up and shutdown switching commands, it can easily lead to inconsistent control rhythms, resulting in: asynchronous response: some devices respond too quickly or too slowly, causing overall system imbalance; overshoot or lag: system oscillation and increased hysteresis time. By identifying these transient coupling paths, rhythm coordination, command sequencing, or beat delay adjustments can be made to make start-up and shutdown control smoother.

[0089] Methods for evaluating the effectiveness of the start-up and shutdown process include:

[0090] A disturbance suppression mechanism is introduced to identify and suppress the instability caused by coal quality disturbances and wind pressure fluctuations to the system in real time during the start-up and shutdown strategy execution, and to evaluate the start-up and shutdown execution effect. The disturbance suppression mechanism includes collecting online coal quality analysis data, wind pressure and wind speed sensor data and coal mill load data to identify the abrupt behavior characteristics of coal particle size, moisture and volatile matter parameters and the dynamic fluctuation characteristics of the air supply system, and to establish a disturbance source feature vector.

[0091] A system disturbance propagation model incorporating control inputs and disturbance inputs is constructed to extract the impact path of disturbances on key output variables of the pulverizing system (including pulverized coal concentration, air-to-pulverized coal ratio, and mill power). Based on the disturbance source feature vector, disturbance suppression strategies are implemented, such as using a pre-feedback adjustment method to correct control command parameters in advance, adding low-pass filtering to the actuator response signal to suppress high-frequency disturbances, and dynamically adjusting the minimum load limit of the mill and the control boundary of the air-to-pulverized coal ratio adjustment zone according to the disturbance intensity.

[0092] The efficiency indicators of the pulverizing system are quantified by constructing an efficiency scoring function, thereby enabling the evaluation of the start-up and shutdown process. The efficiency indicators include response time, energy release rate fit, actuator action consistency, and system stability.

[0093] The preset air-to-powder ratio threshold is set by staff based on historical data analysis results. This historical analysis process includes the system collecting different air-to-powder ratios and calculating their average values ​​as a reference. It can be adjusted by staff during system operation according to the actual situation. Similarly, preset dynamic delay time threshold and minimum loading threshold are set.

[0094] In this embodiment, by constructing a physical-side state vector and combining it with equipment operating states such as mill power, pulverized coal output rate, and air-to-pulverized coal ratio, the energy release rate of each energy operating segment is accurately calculated. By constructing a control subspace for the energy rheology interval, multiple energy operating segments are flexibly divided according to the system's real-time energy release rate and different operating states. Each energy operating segment has a clearly defined energy release rate boundary and control strategy, ensuring that the system can accurately match energy demand at different stages to achieve optimal operating conditions. By monitoring the energy release rate and equipment status in real time, the control objectives of each segment can be flexibly adjusted according to actual operating conditions, thereby achieving optimal energy allocation and utilization.

[0095] By calculating the difference between the current energy release rate and the target rate in real time and introducing a disturbance intensity factor for normalization, the execution delay time of the start-stop control strategy is dynamically adjusted. This mechanism significantly improves the control system's adaptability to fluctuations in operating conditions, making start-stop operations more flexible and controllable, and effectively avoiding control shocks and system oscillations caused by excessively rapid responses. By setting a dynamic delay time threshold, the start-stop strategy switching signal is generated only when the current energy behavior meets the strategy execution conditions, effectively preventing invalid start-stops or unnecessary adjustments, and improving the rationality of control strategy execution and system stability. It can absorb disturbances caused by factors such as coal quality changes and wind pressure fluctuations in real time, achieving active suppression and response optimization of typical disturbance sources, and enhancing the system's robustness to complex operating conditions.

[0096] Example 2

[0097] Please see Figure 2 As shown, parts not described in detail in this embodiment are described in Embodiment 1. An adaptive optimization start-stop control method for a thermal power unit pulverizing system based on APS is provided, including:

[0098] S1. Parse the APS start-stop instructions, extract the process constraint parameters of the pulverizing system, and generate start-stop control targets and load adjustment boundaries in combination with the boiler operating load.

[0099] S2. By collecting operating condition data, extract the operating condition interaction factors that reflect the interaction between coal type, equipment and air-coal interaction, construct a three-dimensional mapping relationship between state, energy and load, and form a physical side state vector;

[0100] S3. Based on the physical side state vector, construct the control subspace of the energy rheological range of the pulverizing system and divide it into different energy operation segments. Calculate the energy release rate of the preset target segment and the current segment, and generate start-stop strategy switching signals.

[0101] S4. Combining the dynamic change curve of APS start-stop command and the current equipment health status data, construct a real-time optimization space for start-stop strategy parameters and perform self-iterative correction of start-stop strategy parameters.

[0102] S5. Receive start / stop strategy switching signals, construct actuator coupling behavior matrix, identify transient coupling paths between multi-source controls; introduce disturbance suppression mechanism, absorb coal quality disturbances and wind pressure fluctuations in real time, and evaluate the effectiveness of start / stop execution process.

[0103] Since the electronic device described in this embodiment is the electronic device used to implement the APS-based adaptive optimization start-stop control system for pulverizing thermal power units in the embodiments of this application, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the APS-based adaptive optimization start-stop control system for pulverizing thermal power units described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the APS-based adaptive optimization start-stop control system for pulverizing thermal power units in the embodiments of this application falls within the scope of protection of this application.

[0104] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0105] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An adaptive optimization start-stop control system for a thermal power unit pulverizing system based on APS, characterized in that, include: The APS instruction parsing module is used to parse APS start-stop instructions, extract process constraint parameters of the pulverizing system, and generate start-stop control targets and load adjustment boundaries in combination with boiler operating load. The operating condition sensing module collects operating condition data, extracts operating condition interaction factors that reflect the interaction between coal type, equipment and air-coal mixture, constructs a three-dimensional mapping relationship between state, energy and load, and forms a physical-side state vector. The segment division module constructs a control subspace for the energy rheological range of the pulverizing system based on the physical side state vector, and divides it into different energy operation segments. It calculates the energy release rate difference between the preset target segment and the current segment, and generates a start-stop strategy switching signal. The method for obtaining the energy operating range includes: Based on the numerical range of the energy release rate, the control subspace of the energy rheology range of the pulverizing system is divided into different energy operation segments; The start-stop correction module combines the dynamic change curve of the APS start-stop command with the current device health status data to construct a real-time optimization space for the start-stop strategy parameters and perform self-iterative correction of the start-stop strategy parameters. The performance evaluation module receives start-stop strategy switching signals, constructs an actuator coupling behavior matrix, identifies transient coupling paths between multi-source controls, introduces a disturbance suppression mechanism, absorbs coal quality disturbances and wind pressure fluctuations in real time, and evaluates the effectiveness of the start-stop execution process.

2. The adaptive optimization start-stop control system for pulverizing thermal power unit based on APS according to claim 1, characterized in that, The method for obtaining the process constraint parameters includes: The system receives structured start-stop commands issued by the APS, performs semantic recognition and parsing on the start-stop commands, and extracts process constraint parameters. The process constraint parameters include minimum air-to-coal ratio, maximum number of start-stop cycles per hour, priority start-stop grouping of coal mills, start-stop response delay time, upper limit of coal mill shell temperature, and lower limit of main steam pressure.

3. The adaptive optimization start-stop control system for pulverizing thermal power units based on APS according to claim 2, characterized in that, The methods for obtaining the start / stop control target and load adjustment boundary include: The APS start-stop command types are analyzed, and the process constraint parameters of the pulverizing system are extracted from them. Combined with the current operating load data of the boiler, a weighted multi-objective optimization function is constructed, and the start-stop control objective is finally output. The load adjustment boundary is derived based on the current operating load state of the boiler and is used to constrain the allowable operating parameter range of the pulverizing system during start-up and shutdown: the initial physical limit load boundary of the equipment is determined according to the design parameters of the core equipment of the pulverizing system; the initial physical limit load boundary is narrowed and corrected based on the safety operation indicators in the process constraint parameters; the corrected initial physical limit load boundary is then verified a second time by combining the minimum stable combustion feed rate and the maximum allowable feed rate corresponding to the current operating load of the boiler, and the final load adjustment boundary is generated.

4. The adaptive optimization start-stop control system for the pulverizing system of a thermal power unit based on APS according to claim 3, characterized in that, The method for obtaining the working condition interaction factor includes: Operating condition data is collected, including online coal quality analysis data, coal mill operating status data, and air pressure and velocity data. The data is preprocessed through standardization and anomaly removal. Based on principal component analysis, operating condition interaction factors reflecting coal type characteristics, equipment response behavior, and the coupling relationship between air and coal conveying are extracted. These operating condition interaction factors include coal type crushing efficiency factor, air-to-coal ratio deviation factor, and equipment dynamic response factor, which characterize the coupling characteristics of the pulverizing system under different operating conditions.

5. The adaptive optimization start-stop control system for the pulverizing system of a thermal power unit based on APS according to claim 4, characterized in that, The method for obtaining the physical side state vector includes: Based on the extracted operating condition interaction factors, coal quality rheological parameters, equipment health status parameters, and gas-solid transport status parameters are further integrated to construct a multi-dimensional state vector describing the operating characteristics of the pulverizing system; and based on this multi-dimensional state vector, state-energy mapping model and state-load mapping model are established respectively. The energy consumption level required per unit load is calculated using the state-energy mapping model, and the maximum load regulation capacity of the system under the current operating conditions is calculated using the state-load mapping model. The output forms the physical-side state vector.

6. The adaptive optimization start-stop control system for the pulverizing system of a thermal power unit based on APS according to claim 5, characterized in that, The method for obtaining the energy operating range also includes: Based on the equipment operating status corresponding to each energy operating segment, the energy release rate of the energy operating segment per unit time is calculated; using the physical side state vector and equipment operating status, a control subspace of the energy rheological range of the pulverizing system is constructed, which reflects the change law and boundary of system energy under different operating conditions.

7. The adaptive optimization start-stop control system for pulverizing thermal power unit based on APS according to claim 6, characterized in that, The method for generating the start / stop strategy switching signal includes: For the target operating segment, the system sets the target energy release rate when the start / stop command is issued. Guide and control the execution direction; preset the energy release rate of the current operating segment per unit time as Calculate the energy release rate difference between the current segment and the target segment during operation; the energy release rate difference is: ; A dynamic delay time function is introduced to calculate the execution delay time of the current start-stop strategy and control the execution rhythm of start-stop actions; The dynamic delay time function is: ;in, Indicates the execution delay time of the current start / stop policy; This indicates the base delay duration of the preset start-stop strategy, which refers to the fixed delay time under no disturbance or baseline conditions, used to ensure a smooth transition of start-stop actions; This represents the perturbation intensity factor, used to adjust the degree of influence of energy deviation on the delay time; A preset dynamic delay time threshold is set. When the execution delay time of the current start / stop strategy is greater than or equal to the preset dynamic delay time threshold, it is determined that the execution conditions are met, and a start / stop strategy switching signal is generated.

8. The adaptive optimization start-stop control system for pulverizing thermal power unit based on APS according to claim 7, characterized in that, The method for constructing the real-time optimization space includes: Collect the historical start-stop command sequence issued by APS, extract the feature information of the historical APS start-stop commands, and form the dynamic change curve of APS start-stop commands; obtain the current equipment health status data and construct the equipment health status vector; Based on the dynamic change curve of APS start / stop commands and the device health status vector, a real-time optimization space for start / stop strategy parameters is constructed. The optimal combination of start / stop strategy parameters is selected in the real-time optimization space using a greedy algorithm, and the start / stop strategy parameters of the current start / stop strategy are iteratively corrected.

9. The adaptive optimization start-stop control system for the pulverizing system of a thermal power unit based on APS according to claim 8, characterized in that, The method for identifying the transient coupling path includes: The control input signals and response output data of each actuator in the pulverizing system are acquired. Based on the response relationship between the input and output of each actuator, an actuator coupling behavior matrix is ​​constructed. The actuator coupling behavior matrix is ​​normalized, and the main coupling path is extracted by the eigenvalue decomposition method. Each actuator is modeled as a node, and the coupling relationship is modeled as an edge. A directed graph of coupling relationship is constructed, and transient coupling paths existing during start-stop switching are identified.

10. The adaptive optimization start-stop control system for the pulverizing system of a thermal power unit based on APS according to claim 9, characterized in that, The methods for evaluating the effectiveness of the start-up and shutdown process include: A disturbance suppression mechanism is introduced to identify and suppress the instability caused by coal quality disturbances and wind pressure fluctuations to the system in real time during the start-up and shutdown strategy execution, and to evaluate the effect of start-up and shutdown execution. The disturbance suppression mechanism includes identifying the abrupt behavior characteristics of coal particle size, moisture, and volatile matter parameters and the dynamic fluctuation characteristics of the air supply system by collecting online coal quality analysis data, wind pressure and wind speed sensor data and coal mill load data, and establishing a disturbance source feature vector. A system disturbance propagation model including control input and disturbance input is constructed to extract the influence path of disturbance on key output variables of the pulverizing system; based on the feature vector of the disturbance source, a disturbance suppression strategy is implemented. By constructing an efficiency scoring function, the efficiency indicators of the pulverizing system are quantified, thereby enabling the evaluation of the start-up and shutdown processes.

Images