Frequency modulation collaborative optimization method and system for coal power unit

Abstract

The application discloses a frequency modulation cooperative optimization method and system for a coal power unit, and belongs to the technical field of operation control of a coal power heating coupling system, and comprises the following steps: under different stable load conditions of the coal power unit, step disturbances are applied to turbine regulating valves and heating system regulating valves, dynamic response data are collected, the heat storage capacity of the coal power unit and the heat storage capacity of the heating system are calculated, and a peak regulation and frequency modulation dynamic model integrating the two is established; based on the model, an integrated auxiliary frequency modulation intelligent control strategy is formulated, an improved genetic algorithm is used to dynamically optimize the load regulation amount distribution scheme of the coordinated system of the coal power unit and the heating system; the control strategy is integrated into the original INFIT optimization control platform of the coal power unit, the optimal coal power unit load instruction and the heating system regulating valve opening degree instruction are calculated and sent in combination with the distribution scheme; and through a closed-loop feedback mechanism based on model predictive control, the operation state is monitored in real time, deviation analysis is performed on the model predictive state, the instruction is dynamically adjusted, and efficient frequency modulation cooperative optimization of the coal power unit is realized.

Description

Technical Field

[0001] This invention belongs to the field of operation and control technology of coal-fired power and heating coupled systems, specifically a frequency regulation and collaborative optimization method and system for coal-fired power units. Background Technology

[0002] In traditional frequency control of coal-fired power units, frequency adjustment typically relies solely on the unit's own heat storage capacity and the regulating system. This approach suffers from slow frequency response and insufficient regulation accuracy when facing rapid fluctuations in grid load. The contradiction between unit load adjustment and heating parameter stability becomes even more pronounced when coal-fired power units are coupled with heating systems, making it difficult for traditional frequency control methods to simultaneously meet both grid frequency regulation requirements and heating stability requirements.

[0003] Currently, although some technologies attempt to coordinate the frequency regulation of heating systems with that of generating units, most of them do not fully consider the energy storage characteristics of heating networks, nor have they established effective frequency regulation models that integrate the energy storage capabilities of generating units and heating systems. This results in severe coupling interference between generating units and heating systems during frequency regulation, affecting the peak-shaving and frequency regulation capabilities of coal-fired power units and the stability of heating parameters. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a frequency regulation collaborative optimization method and system for coal-fired power units. Under different stable load conditions of the coal-fired power unit, step disturbances are applied to the turbine control valves and heating system control valves, dynamic response data is collected, the heat storage capacity of the power unit and the heating system is calculated, and a dynamic model integrating both for peak shaving and frequency regulation is established. Based on this model, an integrated auxiliary frequency regulation intelligent control strategy is formulated, and an improved genetic algorithm is used to dynamically optimize the load adjustment allocation scheme between the unit coordination system and the heating system. The control strategy is integrated into the unit's existing INFIT optimization control platform, and the optimal unit load command and heating system control valve opening command are calculated and sent in conjunction with the allocation scheme. Through a closed-loop feedback mechanism based on model predictive control, the operating status is monitored in real time, and the commands are dynamically adjusted after deviation analysis with the model prediction status, achieving efficient frequency regulation collaborative optimization of the coal-fired power unit.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] Frequency regulation coordination optimization methods for coal-fired power units include:

[0007] Under different stable load conditions of coal-fired power units, step disturbances are applied to the turbine control valve and the heating system control valve respectively. Dynamic response data of unit load, main steam pressure and heating parameters are collected. Based on the dynamic response data, machine learning algorithms are used to calculate the heat storage capacity of the unit and the heat storage capacity of the heating system, and a dynamic model for peak shaving and frequency regulation is established.

[0008] Based on the aforementioned peak-shaving and frequency regulation dynamic model, an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units is formulated. According to the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units, an improved genetic algorithm is used to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system.

[0009] The integrated auxiliary frequency regulation intelligent control strategy of the coal-fired power unit is integrated into the original INFIT optimization control platform of the unit. Through real-time data interaction and combined with the load regulation allocation scheme, the optimal unit load command and heating system valve opening command are calculated. The optimal unit load command is sent to the unit coordination system and the optimal heating system valve opening command is sent to the heating system.

[0010] By using a closed-loop feedback mechanism based on model predictive control, the actual operating status of the unit coordination system and the heating system is monitored in real time. The deviation between the actual operating status and the status predicted by the peak-shaving and frequency regulation dynamic model is analyzed, and the unit load command and the heating system valve opening command are dynamically adjusted according to the deviation.

[0011] Specifically, based on the dynamic response data, the step of using machine learning algorithms to calculate the thermal storage capacity of the heating unit and the thermal storage capacity of the heating system, and establishing a dynamic model for peak shaving and frequency regulation, includes:

[0012] The dynamic response data is preprocessed and then divided into a training set and a validation set.

[0013] The pre-built neural network model is trained using the training set to establish a mapping relationship model between the unit's thermal storage capacity and the heating system's thermal storage capacity and dynamic response data. The model is then validated using the validation set to obtain the trained unit-heating neural network model.

[0014] During unit operation, dynamic response data is collected in real time and input into a trained unit-heating neural network model. The trained unit-heating neural network model calculates the unit's heat storage capacity and the heating system's heat storage capacity based on the input real-time dynamic response data.

[0015] The impact of the unit's thermal storage capacity and the heating system's thermal storage capacity on the unit's peak-shaving and frequency regulation characteristics is analyzed. Based on the impact analysis, the input and output variables of the peak-shaving and frequency regulation dynamic model are determined. The input variables include the unit's thermal storage capacity, the heating system's thermal storage capacity, the grid dispatch command, the unit's current load, the main steam pressure, the heating steam flow rate, and the heating steam temperature. The output variables include the unit's load command and the heating system's valve opening command.

[0016] Based on the physical characteristics of the system and real-time dynamic response data, a mathematical expression for the peak-shaving and frequency-modulation dynamic model is established using a system identification method, thereby generating the peak-shaving and frequency-modulation dynamic model.

[0017] The parameters of the generated peak-shaving and frequency-modulating dynamic model are identified using experimental data.

[0018] Specifically, the process of establishing a mathematical expression for the peak-shaving and frequency-modulation dynamic model based on the system's physical characteristics and real-time dynamic response data, using a system identification method, and generating the peak-shaving and frequency-modulation dynamic model, includes:

[0019] Analyze the physical structure and working principle of coal-fired power units and heating systems, analyze the energy conversion, transfer and storage laws of units and heating systems during peak shaving and frequency regulation, and determine the relationship between system state variables, input variables and output variables;

[0020] By combining real-time dynamic response data, the maximum likelihood estimation method is used to estimate the parameters of the system's transfer function, and a mathematical expression for the peak-shaving and frequency-modulation dynamic model is established.

[0021] The structure and parameters of the established peak-shaving and frequency-modulation dynamic model were verified.

[0022] Specifically, the step of formulating an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units based on the peak-shaving and frequency regulation dynamic model includes:

[0023] The power grid AGC commands are parsed in real time to obtain the power demand and regulation time requirements of the power grid for unit frequency regulation, and these are used as input information for the control strategy.

[0024] Set operational boundary constraints for the unit and heating system;

[0025] Within each control cycle, based on the current operating status of the units and heating system, the grid AGC instructions, and the peak-shaving and frequency regulation dynamic model, the changing trends of unit load, main steam pressure, and heating parameters for multiple future control time steps are predicted.

[0026] With the prediction error minimization as the objective function, and considering the operational boundary constraints of the unit and the heating system, an optimization problem model is constructed. By solving the optimization problem model, the optimal control quantity sequence for multiple future control time steps is obtained, including the unit load command and the heating system valve opening command.

[0027] The first control quantity in the optimal control quantity sequence is used as the control command at the current moment and sent to the unit coordination system and heating system for execution. In the next control cycle, the process is repeated iteratively. Based on the new operating status of the unit and heating system and the grid AGC command, the control strategy is re-formulated to obtain the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units.

[0028] Specifically, the method of dynamically optimizing the load regulation allocation scheme between the unit coordination system and the heating system based on the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units and using an improved genetic algorithm includes:

[0029] The load regulation allocation problem between the unit coordination system and the heating system is transformed into an optimization problem. The overall peak-shaving and frequency regulation performance index of the unit and the heating system is used as the optimization objective, and the operating boundary constraints of the unit and the heating system and the rationality of the load regulation amount are used as the constraints. The overall peak-shaving and frequency regulation performance index of the unit and the heating system includes regulation speed, regulation accuracy and stability.

[0030] An improved genetic algorithm is used to solve the optimization problem; the improved genetic algorithm uses real number encoding, tournament selection as the selection strategy, arithmetic crossover as the crossover operation, and non-uniform mutation as the mutation operation.

[0031] In the iterative process of the genetic algorithm, the fitness of each individual is evaluated, and selection, crossover and mutation operations are performed based on the fitness value. Through iterative evolution, the optimal load adjustment allocation scheme is obtained; the individual refers to a load adjustment allocation scheme.

[0032] Specifically, the process of evaluating the fitness of each individual includes:

[0033] Based on the peak-shaving and frequency regulation dynamic model, the dynamic response of the generating units and heating system within a preset time period is simulated under a given load regulation allocation scheme, and the simulation results are obtained; the dynamic response includes the changes in unit load, main steam pressure and heating parameters;

[0034] The deviation between the simulation results and the grid AGC commands is calculated, as well as the degree of deviation between the operating parameters of the unit and the heating system and the operating boundary constraints. The fitness function is constructed by combining the deviation and the degree of deviation results.

[0035] Substitute the simulation results into the fitness function to calculate the fitness value for each individual.

[0036] Specifically, the integration of the integrated auxiliary frequency regulation intelligent control strategy of the coal-fired power unit into the unit's original INFIT optimized control platform, through real-time data interaction and combined with the load regulation allocation scheme, calculates the optimal unit load command and heating system valve opening command, and sends the optimal unit load command to the unit coordination system and the optimal heating system valve opening command to the heating system, including:

[0037] Analyze the architecture and functions of the original INFIT optimized control platform of the unit to determine the access point of the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units;

[0038] Develop an interface program to add a data interaction interface between the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units and the INFIT optimization control platform. The data to be interacted includes unit load commands, heating system valve opening commands, and actual operating status.

[0039] Based on the current power grid AGC commands, the actual operating status of the generating units and heating systems, and the dynamically optimized load regulation allocation scheme, the optimal generating unit load commands and heating system valve opening commands are calculated.

[0040] Through the data interaction interface, the calculated optimal unit load command is sent to the unit coordination system, and the optimal heating system valve opening command is sent to the heating system. The unit coordination system and the heating system then perform adjustment operations according to the received commands.

[0041] Specifically, the deviation analysis between the actual operating state and the state predicted by the peak-shaving and frequency-regulating dynamic model includes:

[0042] The real-time collected actual operating status data is input into the peak shaving and frequency regulation dynamic model to obtain the state data predicted by the peak shaving and frequency regulation dynamic model.

[0043] Calculate the absolute deviation between the actual operating status data and the predicted status data of the peak-shaving and frequency-regulation dynamic model;

[0044] Analyze the absolute deviation to determine whether it is within the preset allowable range. If the deviation exceeds the allowable range, analyze the cause of the deviation.

[0045] Specifically, the dynamic adjustment of unit load commands and heating system valve opening commands based on deviations includes:

[0046] Based on the results of the deviation analysis, the adjustment amounts of the unit load command and the heating system valve opening command are calculated using the PID control algorithm.

[0047] An incremental adjustment strategy is adopted, which superimposes the adjustment amount with the current unit load command and heating system valve opening command to obtain a new unit load command and heating system valve opening command, and sends them to the unit coordination system and heating system for execution.

[0048] During the adjustment process, the system response after adjustment is monitored in real time, and the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units is optimized and adjusted according to the system response until the deviation between the actual operating state and the model prediction state is within the allowable range.

[0049] The frequency regulation and coordination optimization system for coal-fired power units includes: a data acquisition module, a model building module, a control strategy formulation module, a load regulation allocation module, an instruction module, and a feedback and adjustment module.

[0050] The data acquisition module is used to apply step disturbances to the turbine control valve and the heating system control valve under different stable load conditions of the coal-fired power unit, collect dynamic response data of unit load, main steam pressure and heating parameters, and perform preprocessing.

[0051] The model building module is used to calculate the heat storage capacity of the heating unit and the heat storage capacity of the heating system based on the preprocessed dynamic response data and to establish a dynamic model for peak shaving and frequency regulation using machine learning algorithms.

[0052] The control strategy formulation module is used to formulate an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units based on the peak shaving and frequency regulation dynamic model.

[0053] The load regulation allocation module is used to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system based on the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units and an improved genetic algorithm.

[0054] The instruction module is used to integrate the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units into the original INFIT optimization control platform of the unit. Through real-time data interaction and combined with the load regulation allocation scheme, it calculates the optimal unit load instruction and heating system valve opening instruction, and sends the optimal unit load instruction to the unit coordination system and the optimal heating system valve opening instruction to the heating system.

[0055] The feedback and adjustment module is used to monitor the actual operating status of the unit coordination system and the heating system in real time through a closed-loop feedback mechanism based on model predictive control, and to perform deviation analysis between the actual operating status and the status predicted by the peak-shaving and frequency regulation dynamic model. Based on the deviation, the module dynamically adjusts the unit load command and the heating system valve opening command.

[0056] Compared with the prior art, the beneficial effects of the present invention are:

[0057] 1. This invention proposes a frequency regulation collaborative optimization system for coal-fired power units, and optimizes and improves its architecture, operation steps and processes. The system has the advantages of simple process, low investment and operating costs and low production and working costs.

[0058] 2. This invention proposes a frequency regulation collaborative optimization method for coal-fired power units. By applying step disturbances to coal-fired power units under different stable load conditions to collect data, and using data-driven machine learning algorithms to accurately calculate the heat storage capacity of the unit and the heat storage capacity of the heating system, a dynamic model for peak shaving and frequency regulation integrating the two is established. This helps to more comprehensively and accurately grasp the characteristics of the unit and the heating system, effectively improves the adaptability of the peak shaving and frequency regulation model to actual operating conditions, and makes the model more consistent with the actual operating laws of the unit and the heating system.

[0059] 3. This invention proposes a frequency regulation collaborative optimization method for coal-fired power units. Based on a peak-shaving and frequency regulation model, an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units is formulated. Combined with an improved genetic algorithm, the load regulation allocation scheme is dynamically optimized and integrated into the unit's original INFIT optimization control platform. This method can calculate the optimal unit load command and heating system valve opening command in real time. Through a closed-loop feedback mechanism based on model predictive control, the commands are monitored and dynamically adjusted in real time, realizing collaborative optimization of frequency regulation between the unit and the heating system. This improves the frequency regulation response speed and accuracy of coal-fired power units, enhances the unit's adaptability to the grid's frequency regulation requirements, ensures the safe and stable operation of the grid, and improves the reliability of unit operation. Attached Figure Description

[0060] Figure 1 This is a schematic diagram of the frequency regulation collaborative optimization method for coal-fired power units according to the present invention;

[0061] Figure 2 This is a flowchart illustrating the principle of the frequency regulation collaborative optimization method for coal-fired power units according to the present invention.

[0062] Figure 3 This is a diagram of the frequency regulation collaborative optimization system architecture for coal-fired power units according to the present invention. Detailed Implementation

[0063] Example 1:

[0064] Please see Figure 1 and Figure 2 The present invention provides an embodiment of a frequency regulation coordinated optimization method for coal-fired power units, the method comprising steps S1 to S4, including the following steps:

[0065] S1: Under different stable load conditions of coal-fired power units, step disturbances are applied to the turbine control valve and the heating system control valve respectively. Dynamic response data of unit load, main steam pressure and heating parameters are collected. Based on the dynamic response data, machine learning algorithms are used to calculate the heat storage capacity of the unit and the heat storage capacity of the heating system, and a peak-shaving and frequency regulation dynamic model is established.

[0066] Furthermore, before applying step disturbances to the turbine control valve and the heating system control valve under different stable load conditions of the coal-fired power unit, a comprehensive inspection and commissioning of the coal-fired power unit is also included to ensure that all equipment of the unit operates normally and that the sensors and actuators work reliably; and the data acquisition system is calibrated to ensure that the acquired data is accurate.

[0067] Furthermore, when the coal-fired power unit is at multiple preset stable load conditions, such as 30% rated load, 50% rated load, 70% rated load, and 90% rated load, step disturbances are applied to the turbine control valve and the heating system control valve according to preset step amplitude and step time. The specific time, step amplitude, and step direction of each step disturbance are recorded. The step amplitude of the turbine control valve is determined based on the opening range of the turbine control valve under the current load condition, and the step amplitude of the heating system control valve is determined based on the design parameters and current operating status of the heating system. Step disturbance is a typical load disturbance mode. By suddenly changing the valve opening, the unit's response to this rapid change is observed.

[0068] The amplitude and step time of the step disturbance are determined based on the unit's design parameters and actual operating experience.

[0069] Furthermore, the dynamic response data acquisition of unit load, main steam pressure, and heating parameters further includes: after applying a step disturbance to the turbine control valve and the heating system control valve, using a high-precision data acquisition system to collect heating parameters such as unit load, main steam pressure, heating steam flow rate, heating steam temperature, and heating steam pressure in real time at a preset sampling frequency; and preprocessing the collected data, including removing noise data, filling in missing data, and normalizing the data to ensure the accuracy and consistency of the data.

[0070] S2: Based on the aforementioned peak-shaving and frequency regulation dynamic model, formulate an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units; according to the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units, use an improved genetic algorithm to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system;

[0071] S3: Integrate the integrated auxiliary frequency regulation intelligent control strategy of the coal-fired power unit into the original INFIT optimization control platform of the unit. Through real-time data interaction and combined with the load regulation allocation scheme, calculate the optimal unit load command and heating system valve opening command, and send the optimal unit load command to the unit coordination system and the optimal heating system valve opening command to the heating system.

[0072] Furthermore, the integration of the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units into the unit's existing INFIT optimized control platform specifically involves: analyzing the architecture and functions of the unit's existing INFIT optimized control platform to determine the access point for the integrated auxiliary frequency regulation intelligent control strategy; developing an interface program to realize data interaction between the integrated auxiliary frequency regulation intelligent control strategy and the INFIT optimized control platform, including unit load commands, heating system valve opening commands, and actual operating status; and testing and debugging the INFIT optimized control platform after integrating the integrated auxiliary frequency regulation intelligent control strategy to ensure stable platform operation and collaborative work between various functional modules.

[0073] Furthermore, the step of sending the optimal unit load command to the unit coordination system and the optimal heating system valve opening command to the heating system includes: using the Modbus communication protocol to send the calculated optimal unit load command and heating system valve opening command to the corresponding actuators; during the transmission process, the data is encrypted to ensure data security and integrity; after receiving the commands, the unit coordination system and the heating system perform corresponding adjustment operations according to the command requirements.

[0074] S4: Through a closed-loop feedback mechanism based on model predictive control, the actual operating status of the unit coordination system and the heating system is monitored in real time. The deviation between the actual operating status and the state predicted by the peak-shaving and frequency regulation dynamic model is analyzed, and the unit load command and the heating system valve opening command are dynamically adjusted according to the deviation.

[0075] Furthermore, through a closed-loop feedback mechanism based on model predictive control, the actual operating status of the unit coordination system and the heating system is monitored in real time. This includes: installing sensors at key parts of the unit and the heating system to collect real-time data on actual operating status such as unit load, main steam pressure, heating steam flow rate, heating steam temperature, and heating steam pressure; and using a data acquisition system to transmit the data collected by the sensors to the control center, where the control center processes and analyzes the data in real time.

[0076] Furthermore, the frequency regulation coordinated optimization method for coal-fired power units also includes establishing an evaluation system for the frequency regulation coordinated optimization method, and periodically evaluating the implementation effect of the frequency regulation coordinated optimization method; the evaluation indicators include the frequency regulation performance indicators of the unit, the stability indicators of the heating system, and the economic benefit indicators; based on the evaluation results, the frequency regulation coordinated optimization method is improved and perfected to continuously improve the frequency regulation coordinated optimization level of the unit; the frequency regulation performance indicators of the unit include load response speed and regulation accuracy; the stability indicators of the heating system include the fluctuation range of heating parameters; the economic benefit indicators include fuel consumption and power generation cost.

[0077] Based on the dynamic response data, a machine learning algorithm is used to calculate the thermal storage capacity of the heating unit and the thermal storage capacity of the heating system, and a dynamic model for peak shaving and frequency regulation is established, including:

[0078] S1.1: The dynamic response data is preprocessed and then divided into a training set and a validation set;

[0079] Furthermore, the preprocessing of the dynamic response data includes noise removal, filtering, and time alignment, specifically:

[0080] (1) The noise is removed from the collected dynamic response data by using wavelet transform algorithm. Based on the number of wavelet decomposition layers and the selection of wavelet basis functions, the noise signal is separated from the original signal and the effective dynamic response signal is retained. The wavelet transform algorithm is the prior art in this field and is not an inventive solution of this application. It will not be described in detail here.

[0081] (2) The Kalman filter algorithm is used to filter the dynamic response data after noise removal. Based on the system's state equation and observation equation, the random error in the data is estimated and corrected in real time to improve the smoothness and accuracy of the data. The Kalman filter algorithm is the prior art in this field and is not an inventive solution of this application. It will not be described in detail here.

[0082] (3) Use timestamp information to perform time alignment on the filtered dynamic response data to ensure that the data of different parameters have a consistent time base on the time axis, so as to facilitate subsequent data analysis and modeling.

[0083] S1.2: Use the training set to train the pre-built neural network model, establish a mapping relationship model between the unit's heat storage capacity and the heating system's heat storage capacity and dynamic response data, and use the validation set to validate the model to obtain the trained unit-heating neural network model.

[0084] Furthermore, the specific steps in S1.2 include:

[0085] (1) The pre-built neural network model adopts a multilayer perceptron structure, including an input layer, multiple hidden layers and an output layer. The number of neurons in the input layer is determined according to the dimension of the dynamic response data. The number of neurons in the hidden layer is optimized by cross-validation. The number of neurons in the output layer is 2, which correspond to the heat storage capacity of the unit and the heat storage capacity of the heating system, respectively. The neural network model is the prior art in this field and is not an inventive solution of this application. It will not be described in detail here.

[0086] (2) The backpropagation algorithm is used to train the neural network model. The weights and bias parameters of the neural network are continuously adjusted using the training set data to minimize the error between the model output and the actual value. During the training process, the stochastic gradient descent algorithm is used to update the parameters. The learning rate is dynamically adjusted according to the training process. The backpropagation algorithm is the existing technology in this field and is not an inventive solution of this application. It will not be described in detail here.

[0087] (3) During the training process, the performance of the model is evaluated periodically using the validation set. Based on the error index on the validation set, such as mean square error, the optimal model parameters are selected. When the error on the validation set no longer decreases significantly, the training is stopped, and the trained unit-heating neural network model is obtained.

[0088] S1.3: During unit operation, dynamic response data is collected in real time and input into the trained unit-heating neural network model. The trained unit-heating neural network model calculates the unit's heat storage capacity and the heating system's heat storage capacity based on the input real-time dynamic response data.

[0089] Furthermore, the specific steps in S1.3 include:

[0090] (1) During the operation of the unit, the dynamic response data of the unit load, main steam pressure and heating parameters are collected in real time through the data interface connected to the data acquisition system in real time, according to the same acquisition frequency and parameter settings as the preprocessing stage;

[0091] (2) Perform the same denoising, filtering and time alignment processing on the real-time dynamic response data as in the preprocessing stage to ensure that the data format and features input into the neural network model are consistent with the training set data;

[0092] (3) Input the preprocessed real-time dynamic response data into the trained unit-heating neural network model. The trained unit-heating neural network model quickly calculates the unit heat storage capacity and the heating system heat storage capacity at the current moment based on the mapping relationship learned internally, and outputs the calculation results.

[0093] S1.4: Analyze the impact of the unit's thermal storage capacity and the heating system's thermal storage capacity on the unit's peak-shaving and frequency regulation characteristics. Based on the impact analysis, determine the input and output variables of the peak-shaving and frequency regulation dynamic model. The input variables include the unit's thermal storage capacity, the heating system's thermal storage capacity, the grid dispatch command, the unit's current load, the main steam pressure, the heating steam flow rate, and the heating steam temperature. The output variables include the unit's load command and the heating system's valve opening command.

[0094] Furthermore, the specific steps in S1.4 include:

[0095] (1) Through theoretical analysis and statistical analysis of actual operation data, study the influence of the unit's heat storage capacity and the heat storage capacity of the heating system on the unit's peak-shaving and frequency regulation response speed, regulation range and stability, and establish a quantitative relationship model between the unit's heat storage capacity, the heat storage capacity of the heating system and the unit's peak-shaving and frequency regulation characteristics;

[0096] (2) Based on the analysis results of the quantitative relationship model, the key factors affecting the peak-shaving and frequency regulation characteristics of the unit are determined. The unit's heat storage capacity, the heat storage capacity of the heating system, the grid dispatch instructions, the unit's current load, the main steam pressure, the heating steam flow rate, and the heating steam temperature are used as the input variables of the model. These input variables can fully reflect the operating status of the unit and the heating system as well as the external dispatch requirements.

[0097] (3) Combining the target of peak shaving and frequency regulation of the unit with the actual control requirements, the unit load command and the valve opening command of the heating system are used as the output variables of the model. By controlling these two variables, the coordinated peak shaving and frequency regulation control of the unit and the heating system can be realized.

[0098] S1.5: Based on the physical characteristics of the system and real-time dynamic response data, the mathematical expression of the peak-shaving and frequency-modulation dynamic model is established by using the system identification method, and the peak-shaving and frequency-modulation dynamic model is generated.

[0099] S1.6: Parameter identification of the generated peak-shaving and frequency-modulating dynamic model using experimental data.

[0100] Furthermore, the specific steps in S1.6 include:

[0101] (1) Design an experimental scheme to conduct peak shaving and frequency regulation experiments on coal-fired power units and heating systems under different operating conditions, and record the changes in input variables, output variables and system state variables during the experiment. The input variables include grid dispatch instructions and current unit load, the output variables include unit load instructions and heating system valve opening instructions, and the system state variables include main steam pressure and heating steam flow rate.

[0102] (2) Using experimental data as the basis for parameter identification, the particle swarm optimization algorithm is used to optimize and solve the unknown parameters in the peak-shaving and frequency-modulating dynamic model, so as to minimize the error between the model output and the experimental data. The particle swarm optimization algorithm is the prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0103] (3) During the parameter identification process, the parameters of the particle swarm optimization algorithm are set, such as the population size and the number of iterations, in order to improve the efficiency and accuracy of parameter identification. After multiple iterations of optimization, the optimal model parameters are obtained, and the parameter identification of the peak-shaving and frequency-modulating dynamic model is completed.

[0104] Based on the system's physical characteristics and real-time dynamic response data, a mathematical expression for the peak-shaving and frequency-modulation dynamic model is established using a system identification method, generating the peak-shaving and frequency-modulation dynamic model, including:

[0105] S1.5.1: Analyze the physical structure and working principle of coal-fired power units and heating systems, analyze the energy conversion, transfer and storage laws of units and heating systems during peak shaving and frequency regulation, and determine the relationship between system state variables, input variables and output variables;

[0106] S1.5.2: Combining real-time dynamic response data, the maximum likelihood estimation method is used to estimate the parameters of the system's transfer function and establish a mathematical expression for the peak-shaving and frequency regulation dynamic model; the mathematical expression can accurately describe the dynamic behavior of the unit and heating system during the peak-shaving and frequency regulation process. The maximum likelihood estimation method is existing technology in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0107] S1.5.3: Perform structural and parameter verification on the established peak-shaving and frequency regulation dynamic model. By comparing and analyzing with actual operating data, check the accuracy and reliability of the peak-shaving and frequency regulation dynamic model, and make necessary corrections and optimizations to ensure that the model can truly reflect the peak-shaving and frequency regulation characteristics of the unit and heating system.

[0108] The integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units, based on the aforementioned peak-shaving and frequency regulation dynamic model, includes:

[0109] A1: Real-time parsing of grid AGC commands to obtain the power demand and regulation time requirements of the grid for unit frequency regulation, and using them as input information for control strategies;

[0110] Furthermore, the specific steps of A1 include:

[0111] (1) Receive grid AGC command data in real time through the communication interface with the grid dispatch center. After receiving the data, perform preliminary verification to check the integrity, format correctness and whether it is within a reasonable range. For example, check whether the length of the data packet meets expectations and whether the values ​​in the data are within the upper and lower limits of the unit power regulation.

[0112] (2) According to the format specification of the power grid AGC instruction, the received data is parsed into different fields. Usually, the power grid AGC instruction will include power demand, regulation time requirement, instruction type and priority. The parsing algorithm is used to split the data into various fields according to the definition of the instruction format.

[0113] (3) Extract the specific power value from the parsed power demand field. Since the power value in the power grid AGC instruction may use different units, it is necessary to convert it into a unified unit that the unit control system can recognize according to the actual situation. At the same time, consider the positive and negative signs of the power. A positive sign indicates an increase in power, and a negative sign indicates a decrease in power.

[0114] (4) Extract the specific adjustment time value from the parsed adjustment time requirement field and check whether the time value is within the time range that the unit can respond. If the time value is too short, the unit will not be able to complete the adjustment within the specified time; if the time value is too long, it will affect the frequency stability of the power grid.

[0115] (5) Integrate the extracted and converted power demand and regulation time requirement data to form the input information required for the control strategy. Output the integrated data to the control strategy module in a specific format so that the unit can be frequency controlled according to this information.

[0116] A2: Set operating boundary constraints for the unit and heating system, including the maximum and minimum load limits of the unit, the safe range of the main steam pressure, and the steam flow and temperature limits of the heating system, to ensure that the implementation of the control strategy will not cause the unit and heating system to exceed the safe operating range.

[0117] A3: Within each control cycle, based on the current operating status of the unit and heating system, the grid AGC command, and the peak-shaving and frequency regulation dynamic model, predict the changing trends of unit load, main steam pressure, and heating parameters for multiple future control time steps;

[0118] A4: With the prediction error minimization as the objective function, and considering the operational boundary constraints of the unit and the heating system, an optimization problem model is constructed. By solving this optimization problem model, the optimal control quantity sequence for multiple future control time steps is obtained, including the unit load command and the heating system valve opening command.

[0119] A5: The first control quantity in the optimal control quantity sequence is used as the control command at the current moment and sent to the unit coordination system and heating system for execution. In the next control cycle, the process is repeated iteratively. Based on the new operating status of the unit and heating system and the grid AGC command, the control strategy is re-formulated to obtain the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units.

[0120] The method, based on the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units and employing an improved genetic algorithm to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system, specifically includes:

[0121] B1: The load regulation allocation problem between the unit coordination system and the heating system is transformed into an optimization problem. The overall peak-shaving and frequency regulation performance index of the unit and the heating system is used as the optimization objective, and the operating boundary constraints of the unit and the heating system and the rationality of the load regulation amount are used as the constraint conditions. The overall peak-shaving and frequency regulation performance index of the unit and the heating system includes regulation speed, regulation accuracy and stability.

[0122] B2: An improved genetic algorithm is used to solve the optimization problem; the improved genetic algorithm uses real number encoding, tournament selection as the selection strategy, arithmetic crossover as the crossover operation, and non-uniform mutation as the mutation operation.

[0123] Furthermore, when using an improved genetic algorithm to dynamically optimize the load regulation allocation scheme of the unit coordination system and the heating system, the population size is set to 50-100 individuals, with each individual representing a load regulation allocation scheme; the number of iterations is set to 100-200 generations, adjusted according to the complexity of the problem; in the tournament selection strategy, k individuals are randomly selected from the population each time for comparison, and the individual with the highest fitness is selected to enter the next generation of the population.

[0124] Furthermore, in the arithmetic crossover operation, for two parent individuals, crossover is performed according to a preset crossover probability. The offspring individuals after crossover are obtained by linearly combining the values ​​of the corresponding dimensions of the parent individuals, and the combination coefficient is randomly generated in the interval [0,1]. In the non-uniform mutation operation, for each individual, mutation is performed according to a preset mutation probability. The mutated value is based on the original value with a random perturbation. The magnitude of the random perturbation gradually decreases as the number of iterations increases, so as to achieve the transition from global search to local search. In this invention, a mutation probability of 0.1-0.3 is selected, and a crossover probability of 0.7-0.9 is selected.

[0125] B3: During the iterative process of the genetic algorithm, the fitness of each individual is evaluated, and selection, crossover and mutation operations are performed based on the fitness value. After multiple iterations, the optimal load regulation allocation scheme is obtained. The optimal load regulation allocation scheme can achieve coordinated peak shaving and frequency regulation of the unit and the heating system while meeting the operating constraints of the unit and the heating system, thereby improving the overall peak shaving and frequency regulation performance.

[0126] In this invention, "individual" refers to a load regulation allocation scheme.

[0127] The process of evaluating the fitness of each individual includes:

[0128] B3.1: Based on the peak-shaving and frequency regulation dynamic model, simulate the dynamic response of the unit and heating system within a preset time period under a given load regulation allocation scheme, and obtain the simulation results; the dynamic response includes the changes in unit load, main steam pressure and heating parameters, and the preset time period refers to a future period;

[0129] B3.2: Calculate the deviation between the simulation results and the grid AGC commands, as well as the degree of deviation between the operating parameters of the unit and the heating system and the operating boundary constraints. Combine the deviation and degree of deviation results to construct the fitness function.

[0130] Furthermore, the specific steps in B3.2 include:

[0131] (1) Obtain simulation results;

[0132] (2) Obtain AGC command data from the power grid dispatch center and clarify the power demand command values ​​at each time point;

[0133] (3) Obtain the operating boundary constraints of the unit and the heating system;

[0134] (4) For each time point, calculate the difference between the unit power in the simulation results and the power demand command value in the grid AGC command, i.e., the deviation value;

[0135] (5) For the unit power, calculate the degree of deviation from the upper and lower power limits; the degree of deviation from the upper power limit is the difference between the unit power at the i-th time point in the simulation results and the upper power limit, and then divided by the upper power limit; the degree of deviation from the lower power limit is the difference between the lower power limit and the unit power at the i-th time point in the simulation results, and then divided by the lower power limit.

[0136] (6) For the parameters of the heating system, the deviation degree of the heating system parameters is obtained by using the deviation degree calculation method in (5) based on the minimum and maximum values ​​of their safe range;

[0137] (7) Summarize the degree of deviation at all time points and calculate the average degree of deviation;

[0138] (8) The fitness value is obtained by weighted summation of the average deviation and the deviation value.

[0139] B3.3: Substitute the simulation results into the fitness function to calculate the fitness value of each individual. The higher the fitness value, the better the load regulation allocation scheme is, and the better it can meet the frequency regulation requirements of the power grid and the operation requirements of the unit and heating system.

[0140] The process of integrating the integrated auxiliary frequency regulation intelligent control strategy of the coal-fired power unit into the unit's original INFIT optimized control platform, through real-time data interaction and combined with the load regulation allocation scheme, calculates the optimal unit load command and heating system valve opening command, and sends the optimal unit load command to the unit coordination system and the optimal heating system valve opening command to the heating system, including:

[0141] S3.1: Analyze the architecture and functions of the original INFIT optimized control platform of the unit, and determine the access point of the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units;

[0142] S3.2: Develop an interface program to add a data interaction interface between the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units and the INFIT optimization control platform. The data to be interacted includes unit load commands, heating system valve opening commands, and actual operating status.

[0143] S3.3: Based on the current power grid AGC instructions, the actual operating status of the units and heating system, and the dynamically optimized load regulation allocation scheme, calculate the optimal unit load instructions and heating system valve opening instructions.

[0144] S3.4: Through the data interaction interface, the optimal unit load command calculated is sent to the unit coordination system, and the optimal heating system valve opening command is sent to the heating system. The unit coordination system and the heating system perform corresponding adjustment operations according to the received commands to realize coordinated peak shaving and frequency regulation control of the unit and the heating system.

[0145] The deviation analysis between the actual operating state and the state predicted by the peak-shaving and frequency-modulation dynamic model includes:

[0146] S4.1: Input the real-time collected actual operating status data into the peak shaving and frequency regulation dynamic model to obtain the status data predicted by the peak shaving and frequency regulation dynamic model;

[0147] S4.2: Calculate the absolute deviation between the actual operating status data and the predicted status data of the peak-shaving and frequency-regulating dynamic model;

[0148] S4.3: Analyze the absolute deviation to determine whether the deviation is within the preset allowable range. If the deviation exceeds the allowable range, further analyze the cause of the deviation.

[0149] The dynamic adjustment of unit load commands and heating system valve opening commands based on deviations includes:

[0150] C1: Based on the results of the deviation analysis, the adjustment amount of the unit load command and the valve opening command of the heating system is calculated using the PID control algorithm;

[0151] C2: An incremental adjustment strategy is adopted, which superimposes the adjustment amount with the current unit load command and heating system valve opening command to obtain a new unit load command and heating system valve opening command, and sends them to the unit coordination system and heating system for execution.

[0152] C3: During the adjustment process, monitor the system response after adjustment in real time, and further optimize and adjust the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units based on the system response until the deviation between the actual operating state and the model prediction state is within the allowable range.

[0153] Example 2:

[0154] Please see Figure 3 Another embodiment of the present invention provides: a frequency regulation collaborative optimization system for coal-fired power units, comprising:

[0155] Data acquisition module, model building module, control strategy formulation module, load regulation allocation module, instruction module, feedback and adjustment module;

[0156] The data acquisition module is used to apply step disturbances to the turbine control valves and heating system control valves under different stable load conditions of the coal-fired power unit, collect dynamic response data of unit load, main steam pressure and heating parameters, and perform preprocessing to provide high-quality data for subsequent analysis.

[0157] The model building module is used to calculate the heat storage capacity of the heating unit and the heat storage capacity of the heating system based on the preprocessed dynamic response data and to establish a dynamic model for peak shaving and frequency regulation.

[0158] The control strategy formulation module is used to formulate an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units based on the peak shaving and frequency regulation dynamic model.

[0159] The load regulation allocation module is used to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system based on the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units and an improved genetic algorithm.

[0160] The instruction module is used to integrate the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units into the original INFIT optimization control platform of the unit. Through real-time data interaction and combined with the load regulation allocation scheme, it calculates the optimal unit load instruction and heating system valve opening instruction, and sends the optimal unit load instruction to the unit coordination system and the optimal heating system valve opening instruction to the heating system.

[0161] The feedback and adjustment module is used to monitor the actual operating status of the unit coordination system and the heating system in real time through a closed-loop feedback mechanism based on model predictive control. It analyzes the deviation between the actual operating status and the status predicted by the peak-shaving and frequency regulation dynamic model, and dynamically adjusts the unit load command and the heating system valve opening command according to the deviation.

[0162] The data acquisition module includes: a disturbance application unit, a data acquisition unit, and a preprocessing unit;

[0163] The disturbance application unit is used to apply step disturbances to the turbine control valve and the heating system control valve respectively through a preset step disturbance generator when the coal-fired power unit is under multiple preset stable load conditions. The amplitude and duration of the step disturbance are determined according to the design parameters of the unit and actual operating experience to stimulate the dynamic response of the unit and the heating system.

[0164] The data acquisition unit is used to acquire dynamic response data of heating parameters such as unit load, main steam pressure, steam flow, steam temperature, and return water temperature in the heating system in real time while applying a step disturbance using a high-precision data acquisition system.

[0165] The preprocessing unit is used to perform preliminary preprocessing on the acquired dynamic response data, including noise removal, filtering, and time alignment.

[0166] The model building module includes: a data partitioning unit, a capability calculation unit, a variable determination unit, and a model building unit;

[0167] The data partitioning unit is used to divide the preprocessed dynamic response data into training and validation sets.

[0168] The capacity calculation unit is used to collect dynamic response data in real time during unit operation and perform the same noise reduction, filtering and time alignment processing as in the preprocessing stage. The preprocessed real-time dynamic response data is input into the trained unit-heating neural network model. The unit-heating neural network model quickly calculates the unit's heat storage capacity and the heating system's heat storage capacity at the current moment based on the mapping relationship it has learned internally, and outputs the calculation results.

[0169] The variable determination unit is used to analyze the impact of the unit's thermal storage capacity and the heating system's thermal storage capacity on the unit's peak-shaving and frequency regulation characteristics. Based on the impact analysis, the input and output variables of the model are determined.

[0170] The model building unit is used to establish the mathematical expression of the peak-shaving and frequency-modulation dynamic model based on the physical characteristics of the system and real-time dynamic response data, using system identification methods.

[0171] The control strategy formulation module includes: an instruction parsing unit, a constraint consideration unit, and a strategy formulation unit;

[0172] The instruction parsing unit is used to parse the grid AGC instructions in real time, obtain the power demand and regulation time requirements of the grid for unit frequency regulation, and use them as important input information for the control strategy.

[0173] The constraint consideration unit is used to comprehensively consider the operating boundary constraints of the unit and the heating system, including the maximum and minimum load limits of the unit, the safe range of the main steam pressure, and the steam flow and temperature limits of the heating system, to ensure that the implementation of the control strategy will not cause the unit and the heating system to exceed the safe operating range.

[0174] The strategy formulation unit is used to formulate an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units based on the peak-shaving and frequency regulation dynamic model, using model predictive control methods, and according to the grid AGC instructions and the operating boundary constraints of the unit and heating system.

[0175] The load regulation allocation module includes: a problem transformation unit, an algorithm optimization unit, a fitness evaluation unit, and a scheme generation unit;

[0176] The problem transformation unit is used to transform the load regulation allocation problem between the unit coordination system and the heating system into an optimization problem.

[0177] The algorithm optimization unit is used to solve the optimization problem using an improved genetic algorithm.

[0178] The fitness evaluation unit is used to simulate the dynamic response of the unit and heating system in the future under a given load regulation allocation scheme based on the peak shaving and frequency regulation dynamic model, and to calculate the deviation between the simulation results and the grid AGC command, as well as the degree of deviation between the operating parameters of the unit and heating system and the operating boundary constraints, and to construct the fitness function.

[0179] The scheme generation unit is used to evaluate the fitness of each individual during the iterative process of the genetic algorithm, and after multiple iterations, obtain the optimal load adjustment allocation scheme.

[0180] The instruction module includes: an interface modification unit, an instruction calculation unit, and an instruction sending unit;

[0181] The interface modification unit is used to modify the interface of the original INFIT optimized control platform of the unit, add a data interaction interface with the integrated auxiliary frequency regulation intelligent control strategy of the coal-fired power unit, and realize real-time data transmission between the two.

[0182] The instruction calculation unit is used to calculate the optimal unit load instruction and heating system valve opening instruction in the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units, based on the current power grid AGC instruction, the actual operating status of the unit and heating system, and the dynamically optimized load regulation allocation scheme.

[0183] The instruction sending unit is used to send the calculated optimal unit load instruction to the unit coordination system and the optimal heating system valve opening instruction to the heating system through the data interaction interface. The unit coordination system and the heating system perform corresponding adjustment operations according to the received instructions to realize the coordinated peak shaving and frequency regulation control of the unit and the heating system.

[0184] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the present invention. All of these variations are within the protection scope of the present invention.

Claims

1. A frequency regulation coordination optimization method for coal-fired power units, characterized in that, include: Under different stable load conditions of coal-fired power units, step disturbances are applied to the turbine control valve and the heating system control valve respectively. Dynamic response data of unit load, main steam pressure and heating parameters are collected. Based on the dynamic response data, machine learning algorithms are used to calculate the heat storage capacity of the unit and the heat storage capacity of the heating system, and a dynamic model for peak shaving and frequency regulation is established. Based on the aforementioned peak-shaving and frequency regulation dynamic model, an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units is formulated. According to the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units, an improved genetic algorithm is used to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system. The integrated auxiliary frequency regulation intelligent control strategy of the coal-fired power unit is integrated into the original INFIT optimization control platform of the unit. Through real-time data interaction and combined with the load regulation allocation scheme, the optimal unit load command and heating system valve opening command are calculated. The optimal unit load command is sent to the unit coordination system and the optimal heating system valve opening command is sent to the heating system. Through a closed-loop feedback mechanism based on model predictive control, the actual operating status of the unit coordination system and the heating system is monitored in real time. The deviation between the actual operating status and the state predicted by the peak-shaving and frequency regulation dynamic model is analyzed, and the unit load command and the heating system valve opening command are dynamically adjusted according to the deviation. Based on the dynamic response data, a machine learning algorithm is used to calculate the thermal storage capacity of the heating unit and the thermal storage capacity of the heating system, and a dynamic model for peak shaving and frequency regulation is established, including: The dynamic response data is preprocessed and then divided into a training set and a validation set. The pre-built neural network model is trained using the training set to establish a mapping relationship model between the unit's thermal storage capacity and the heating system's thermal storage capacity and dynamic response data. The model is then validated using the validation set to obtain the trained unit-heating neural network model. During unit operation, dynamic response data is collected in real time and input into a trained unit-heating neural network model. The trained unit-heating neural network model calculates the unit's heat storage capacity and the heating system's heat storage capacity based on the input real-time dynamic response data. The impact of the unit's thermal storage capacity and the heating system's thermal storage capacity on the unit's peak-shaving and frequency regulation characteristics is analyzed. Based on the impact analysis, the input and output variables of the peak-shaving and frequency regulation dynamic model are determined. The input variables include the unit's thermal storage capacity, the heating system's thermal storage capacity, the grid dispatch command, the unit's current load, the main steam pressure, the heating steam flow rate, and the heating steam temperature. The output variables include the unit's load command and the heating system's valve opening command. Based on the physical characteristics of the system and real-time dynamic response data, a mathematical expression for the peak-shaving and frequency-modulation dynamic model is established using the system identification method, thereby generating the peak-shaving and frequency-modulation dynamic model. The parameters of the generated peak-shaving and frequency-modulating dynamic model are identified using experimental data. The method, based on the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units and employing an improved genetic algorithm to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system, includes: The load regulation allocation problem between the unit coordination system and the heating system is transformed into an optimization problem. The overall peak-shaving and frequency regulation performance index of the unit and the heating system is used as the optimization objective, and the operating boundary constraints of the unit and the heating system and the rationality of the load regulation amount are used as the constraints. The overall peak-shaving and frequency regulation performance index of the unit and the heating system includes regulation speed, regulation accuracy and stability. An improved genetic algorithm is used to solve the optimization problem; the improved genetic algorithm uses real number encoding, tournament selection as the selection strategy, arithmetic crossover as the crossover operation, and non-uniform mutation as the mutation operation. In the iterative process of the genetic algorithm, the fitness of each individual is evaluated, and selection, crossover and mutation operations are performed based on the fitness value. Through iterative evolution, the optimal load adjustment allocation scheme is obtained; the individual refers to a load adjustment allocation scheme.

2. The frequency regulation and coordinated optimization method for coal-fired power units as described in claim 1, characterized in that, The mathematical expression for the peak-shaving and frequency-modulation dynamic model is established using a system identification method based on the system's physical characteristics and real-time dynamic response data, generating the peak-shaving and frequency-modulation dynamic model, including: Analyze the physical structure and working principle of coal-fired power units and heating systems, analyze the energy conversion, transfer and storage laws of units and heating systems during peak shaving and frequency regulation, and determine the relationship between system state variables, input variables and output variables; By combining real-time dynamic response data, the maximum likelihood estimation method is used to estimate the parameters of the system's transfer function, and a mathematical expression for the peak-shaving and frequency-modulation dynamic model is established. The structure and parameters of the established peak-shaving and frequency-modulation dynamic model were verified.

3. The frequency regulation coordination optimization method for coal-fired power units as described in claim 2, characterized in that, The integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units, based on the aforementioned peak-shaving and frequency regulation dynamic model, includes: The power grid AGC commands are parsed in real time to obtain the power demand and regulation time requirements of the power grid for unit frequency regulation, and these are used as input information for the control strategy. Set operational boundary constraints for the unit and heating system; Within each control cycle, based on the current operating status of the units and heating system, the grid AGC instructions, and the peak-shaving and frequency regulation dynamic model, the changing trends of unit load, main steam pressure, and heating parameters for multiple future control time steps are predicted. With the prediction error minimization as the objective function, and considering the operational boundary constraints of the unit and the heating system, an optimization problem model is constructed. By solving the optimization problem model, the optimal control quantity sequence for multiple future control time steps is obtained, including the unit load command and the heating system valve opening command. The first control quantity in the optimal control quantity sequence is used as the control command at the current moment and sent to the unit coordination system and heating system for execution. In the next control cycle, the process is repeated iteratively. Based on the new operating status of the unit and heating system and the grid AGC command, the control strategy is re-formulated to obtain the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units.

4. The frequency regulation coordination optimization method for coal-fired power units as described in claim 3, characterized in that, The process of evaluating the fitness of each individual includes: Based on the peak-shaving and frequency regulation dynamic model, the dynamic response of the generating units and heating system within a preset time period is simulated under a given load regulation allocation scheme, and the simulation results are obtained; the dynamic response includes the changes in unit load, main steam pressure and heating parameters; The deviation between the simulation results and the grid AGC commands is calculated, as well as the degree of deviation between the operating parameters of the unit and the heating system and the operating boundary constraints. The fitness function is constructed by combining the deviation and the degree of deviation results. Substitute the simulation results into the fitness function to calculate the fitness value for each individual.

5. The frequency regulation coordination optimization method for coal-fired power units as described in claim 4, characterized in that, The process of integrating the integrated auxiliary frequency regulation intelligent control strategy of the coal-fired power unit into the unit's original INFIT optimized control platform, through real-time data interaction and combined with the load regulation allocation scheme, calculates the optimal unit load command and heating system valve opening command, and sends the optimal unit load command to the unit coordination system and the optimal heating system valve opening command to the heating system, including: Analyze the architecture and functions of the original INFIT optimized control platform of the unit to determine the access point of the integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units; Develop an interface program to add a data interaction interface between the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units and the INFIT optimization control platform. The data to be interacted includes unit load commands, heating system valve opening commands, and actual operating status. Based on the current power grid AGC commands, the actual operating status of the generating units and heating systems, and the dynamically optimized load regulation allocation scheme, the optimal generating unit load commands and heating system valve opening commands are calculated. Through the data interaction interface, the optimal unit load command calculated is sent to the unit coordination system, and the optimal heating system valve opening command is sent to the heating system. The unit coordination system and the heating system then perform adjustment operations according to the received commands.

6. The frequency regulation coordinated optimization method for coal-fired power units as described in claim 5, characterized in that, The deviation analysis between the actual operating state and the state predicted by the peak-shaving and frequency-modulation dynamic model includes: The real-time collected actual operating status data is input into the peak shaving and frequency regulation dynamic model to obtain the status data predicted by the peak shaving and frequency regulation dynamic model. Calculate the absolute deviation between the actual operating status data and the predicted status data of the peak-shaving and frequency-regulation dynamic model; Analyze the absolute deviation to determine whether it is within the preset allowable range. If the deviation exceeds the allowable range, analyze the cause of the deviation.

7. The frequency regulation coordination optimization method for coal-fired power units as described in claim 6, characterized in that, The dynamic adjustment of unit load commands and heating system valve opening commands based on deviations includes: Based on the results of the deviation analysis, the adjustment amount of the unit load command and the valve opening command of the heating system is calculated using the PID control algorithm; An incremental adjustment strategy is adopted, which superimposes the adjustment amount with the current unit load command and heating system valve opening command to obtain a new unit load command and heating system valve opening command, and sends them to the unit coordination system and heating system for execution. During the adjustment process, the system response after adjustment is monitored in real time, and the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units is optimized and adjusted according to the system response until the deviation between the actual operating state and the model prediction state is within the allowable range.

8. A frequency regulation collaborative optimization system for coal-fired power units, used to implement the frequency regulation collaborative optimization method for coal-fired power units as described in any one of claims 1-7, characterized in that, include: Data acquisition module, model building module, control strategy formulation module, load regulation allocation module, instruction module, feedback and adjustment module; The data acquisition module is used to apply step disturbances to the turbine control valve and the heating system control valve under different stable load conditions of the coal-fired power unit, collect dynamic response data of unit load, main steam pressure and heating parameters, and perform preprocessing. The model building module is used to calculate the heat storage capacity of the heating unit and the heat storage capacity of the heating system based on the preprocessed dynamic response data and to establish a dynamic model for peak shaving and frequency regulation using machine learning algorithms. The control strategy formulation module is used to formulate an integrated auxiliary frequency regulation intelligent control strategy for coal-fired power units based on the peak shaving and frequency regulation dynamic model. The load regulation allocation module is used to dynamically optimize the load regulation allocation scheme between the unit coordination system and the heating system based on the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units and an improved genetic algorithm. The instruction module is used to integrate the integrated auxiliary frequency regulation intelligent control strategy of coal-fired power units into the original INFIT optimization control platform of the unit. Through real-time data interaction and combined with the load regulation allocation scheme, it calculates the optimal unit load instruction and heating system valve opening instruction, and sends the optimal unit load instruction to the unit coordination system and the optimal heating system valve opening instruction to the heating system. The feedback and adjustment module is used to monitor the actual operating status of the unit coordination system and the heating system in real time through a closed-loop feedback mechanism based on model predictive control, and to perform deviation analysis between the actual operating status and the status predicted by the peak-shaving and frequency regulation dynamic model. Based on the deviation, the module dynamically adjusts the unit load command and the heating system valve opening command.

Images