A method and apparatus for optimizing frequency regulation parameters of coal-fired power units
By acquiring data on the primary frequency regulation actions of coal-fired power units to quantitatively evaluate the effect and dynamically update the parameter model, the problem of frequency regulation parameters of coal-fired power units being unable to adapt to the drift of operating characteristics has been solved, thereby improving frequency regulation performance and grid stability.
Patent Information
- Application Number
- CN202610981664.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-02
- Publication Date
- 2026-11-17
AI Technical Summary
Fixed values of primary frequency regulation parameters for coal-fired power units are difficult to adapt to the drift in operating characteristics caused by factors such as equipment aging, fuel changes, and seasonal changes, resulting in a decline in frequency regulation performance, an inability to effectively suppress grid frequency fluctuations, and an impact on the safe and stable operation of the power grid.
By acquiring the operating condition data and action data of a single frequency modulation operation, the frequency modulation effect score is quantitatively evaluated, the parameter optimization process is automatically triggered, the parameter optimization model is dynamically updated, and a target frequency modulation parameter value combination that adapts to the current operating state is generated.
It improves the regulation accuracy and response performance of primary frequency regulation of coal-fired power units, effectively suppresses grid frequency fluctuations, and provides a guarantee for the safe and stable operation of the power grid.
Smart Images

Figure CN123121648A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of coal-fired power unit technology, and in particular to a method and apparatus for optimizing frequency regulation parameters of coal-fired power units. Background Technology
[0002] Primary frequency regulation of coal-fired power units is the first line of defense to ensure grid frequency stability when the grid frequency deviates from its rated value. Specifically, it refers to the automatic adjustment of active power output by the coal-fired power unit control system based on the frequency deviation, quickly smoothing out grid frequency fluctuations. The unit achieves power regulation by adjusting the opening of the speed regulating valve and changing the steam intake. This process relies on core frequency regulation parameters such as speed inequality and the actuation of the primary frequency regulation valve. Therefore, the performance of the unit's primary frequency regulation largely depends on the adaptability of the frequency regulation parameter values.
[0003] Currently, frequency regulation parameters are mostly fixed in advance by technicians based on simulation calculations, limited field tests, and historical operating experience. However, during the actual operation of coal-fired power units, their operating characteristics will slowly drift due to factors such as equipment aging, fuel changes, and seasonal changes. This makes it difficult for the pre-set fixed frequency regulation parameters to maintain optimal primary frequency regulation performance. This may result in insufficient regulation accuracy of coal-fired power units, making it impossible to effectively smooth grid frequency fluctuations and affecting the safe and stable operation of the power grid. Summary of the Invention
[0004] In view of the above problems, the present invention provides a method and apparatus for optimizing the frequency regulation parameters of a coal-fired power unit, the main purpose of which is to improve the regulation accuracy of the primary frequency regulation of the coal-fired power unit.
[0005] To solve the above-mentioned technical problems, the present invention proposes the following solution: In a first aspect, the present invention provides a method for optimizing frequency regulation parameter values of a coal-fired power unit, the method comprising: In response to the coal-fired power unit completing a frequency regulation action based on the combination of frequency regulation parameter values, the frequency regulation condition data and frequency regulation action data during the frequency regulation action are obtained. The frequency regulation action data includes the combination of frequency regulation parameter values between various frequency regulation parameters and the actual response data generated by the coal-fired power unit based on the combination of frequency regulation parameter values during the frequency regulation action. Based on the frequency regulation operating condition data and the frequency regulation action data, determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions; Based on the frequency modulation effect score, determine whether it is necessary to initiate the frequency modulation parameter value optimization process; If it is determined that activation is required, the pre-trained parameter value optimization model is updated and trained based on the frequency modulation operating condition data, the frequency modulation action data, and the frequency modulation effect score to obtain the updated parameter value optimization model. Based on the updated parameter value optimization model, a target frequency modulation parameter value combination is generated for subsequent frequency modulation actions.
[0006] Secondly, the present invention provides a device for optimizing frequency regulation parameters of a coal-fired power unit, the device comprising: The acquisition unit is used to respond to the coal-fired power unit completing a frequency regulation action based on the combination of frequency regulation parameter values, and to acquire frequency regulation condition data and frequency regulation action data during the frequency regulation action. The frequency regulation action data includes the combination of frequency regulation parameter values between various frequency regulation parameters and the actual response data generated by the coal-fired power unit based on the combination of frequency regulation parameter values during the frequency regulation action. The scoring determination unit is used to determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions based on the frequency regulation condition data and frequency regulation action data obtained by the acquisition unit. An optimization judgment unit is used to determine whether the frequency modulation effect score determined by the score determination unit needs to be started, and whether the frequency modulation parameter value optimization process needs to be initiated. The model training unit is used to update and train the pre-trained parameter value optimization model based on the frequency modulation operating condition data, the frequency modulation action data, and the frequency modulation effect score if the optimization judgment unit determines that it needs to be started, so as to obtain the updated parameter value optimization model. The numerical optimization unit is used to optimize the model based on the updated parameter values obtained by the model training unit, and generate a combination of target frequency modulation parameter values for subsequent frequency modulation actions.
[0007] To achieve the above objectives, according to a third aspect of the present invention, a storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device where the storage medium is located is controlled to execute the frequency regulation parameter value optimization method of the coal-fired power unit described in the first aspect.
[0008] To achieve the above objectives, according to a fourth aspect of the present invention, a processor is provided for running a program, wherein the program executes the frequency regulation parameter value optimization method for coal-fired power units described in the first aspect.
[0009] By employing the above technical solution, this invention provides a method and apparatus for optimizing frequency regulation parameters of a coal-fired power unit. This method acquires frequency regulation operating data and frequency regulation action data by responding to a frequency regulation action performed by the coal-fired power unit based on a combination of frequency regulation parameters. Subsequently, based on the frequency regulation operating data and frequency regulation action data, a frequency regulation effect score is determined for this frequency regulation action, achieving a quantitative evaluation of the unit's single frequency regulation performance. This allows for a direct assessment of the suitability of the current combination of frequency regulation parameters to the actual operating state of the unit. Furthermore, the quantitative effect score determines whether the frequency regulation parameter optimization process needs to be initiated, enabling the optimization action to be automatically triggered based on actual performance, avoiding the lag and subjectivity of parameter adjustments. Once optimization is initiated, the frequency regulation operating data, frequency regulation action data, and frequency regulation effect score are used to update and train a pre-trained parameter value optimization model. This dynamic update mechanism overcomes the technical defect that fixed frequency regulation parameter values cannot follow changes in unit characteristics, leading to a decline in frequency regulation performance. Finally, based on the updated parameter value optimization model, a combination of target frequency regulation parameter values is generated for subsequent frequency regulation actions. The generated parameter values can better adapt to the current actual operating status of the unit, which is conducive to improving the regulation accuracy and response performance of the unit's primary frequency regulation, and thus more effectively suppressing grid frequency fluctuations, providing a guarantee for the safe and stable operation of the grid.
[0010] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description
[0011] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This invention provides a flowchart of a method for optimizing frequency regulation parameters of a coal-fired power unit according to an embodiment of the present invention. Figure 2 This invention provides a flowchart of another method for optimizing frequency regulation parameters of a coal-fired power unit according to an embodiment of the present invention. Figure 3 This diagram shows a block diagram of a frequency regulation parameter optimization device for a coal-fired power unit according to an embodiment of the present invention. Figure 4 This invention provides a block diagram of another frequency regulation parameter value optimization device for a coal-fired power unit. Detailed Implementation
[0012] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
[0013] In existing technologies, the primary frequency regulation parameters of coal-fired power units are mainly pre-set through manual offline simulation, limited tests, and historical experience. However, the actual operating characteristics of the unit will slowly drift with factors such as equipment aging, fuel changes, and seasonal changes, making it difficult to maintain the fixed frequency regulation parameters at their optimal level. This leads to deterioration in frequency regulation performance, insufficient regulation accuracy, inability to effectively suppress grid frequency fluctuations, and even threats to grid operation safety. Furthermore, manual optimization methods are subject to lag and subjectivity, and cannot achieve continuous adaptive optimization of parameters.
[0014] To address the aforementioned issues, this invention uses real data from a single, complete frequency regulation operation of a coal-fired power unit as its foundation. It employs a quantitative evaluation of frequency regulation performance to intelligently trigger an optimization process—dynamically updating the parameter optimization model using newly generated data and generating target frequency regulation parameter values for subsequent operations. The entire process, through a closed-loop iterative flow of "data acquisition—performance evaluation—trigger optimization—model update—parameter optimization," achieves continuous adaptive optimization of the real-time characteristics of randomized frequency regulation parameters, fundamentally solving the technical challenge of fixed parameters failing to follow the drift of unit characteristics.
[0015] Next, combined Figure 1 This invention provides a method for optimizing frequency regulation parameters of a coal-fired power unit, and its specific execution steps are as follows: Figure 1 As shown, it includes: 101. In response to the coal-fired unit completing a frequency regulation action based on the combination of frequency regulation parameter values, acquire the frequency regulation condition data and frequency regulation action data during the first frequency regulation action.
[0016] 102. Based on the frequency regulation operating condition data and frequency regulation action data, determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions.
[0017] The frequency regulation action data includes the combination of frequency regulation parameter values between various frequency regulation parameters and the actual response data generated by the coal-fired unit during the frequency regulation action based on the combination of frequency regulation parameter values.
[0018] In this invention, the control system of the coal-fired unit monitors the power grid frequency in real time. Once the frequency fluctuation exceeds the preset frequency regulation dead zone, a frequency regulation event is immediately initiated, and data acquisition is carried out simultaneously. This acquisition process continues until the power grid frequency returns to the normal range and the unit's active power stabilizes at the new load point. During this period, frequency regulation condition data and frequency regulation action data are acquired simultaneously: the frequency regulation condition data includes the overall operating background of the unit at the time of the frequency regulation, including static snapshot information such as the reference unit load, reference main steam pressure, and initial comprehensive valve opening at the start of the frequency regulation; the frequency regulation action data fully records the entire process of the frequency regulation event, including dynamic process data such as the actual load response curve, the adjustment time calculated based on the curve, the actual main steam pressure curve, and the actual comprehensive valve opening curve, as well as the parameter value combination actually used in this frequency regulation (such as the speed inequality setpoint, frequency regulation direction, target frequency regulation load command, and other command input data). Subsequently, in step 102, the frequency regulation effect score of the coal-fired unit during this frequency regulation execution process will be determined based on the aforementioned frequency regulation operating condition data and frequency regulation action data.
[0019] It should be noted that the frequency dead zone (also known as the speed dead zone) is a pre-set frequency non-operation range to avoid the unit from frequently starting and stopping frequency regulation due to small frequency fluctuations in the power grid, which would lead to excessive wear of the speed control valve and instability in the steady-state operation of the unit.
[0020] In step 102, the frequency regulation speed score, frequency regulation accuracy score, frequency regulation safety score, and frequency regulation economy score can be determined based on the benchmark unit load, benchmark main steam pressure, initial comprehensive valve position opening, frequency difference, speed unequal rate setpoint, actual load response curve, adjustment time obtained from the actual load response curve, actual main steam pressure curve, and actual comprehensive valve position opening curve. Then, the weighted average score among the frequency regulation speed score, frequency regulation accuracy score, frequency regulation safety score, and frequency regulation economy score is determined as the frequency regulation effect score.
[0021] 103. Based on the frequency modulation effect score, determine whether it is necessary to start the frequency modulation parameter value optimization process.
[0022] In this step, the frequency modulation parameter value optimization process can be initiated using two different decision-making logics: The first type is the single-trigger judgment logic: when the frequency modulation effect score corresponding to the completed frequency modulation action is lower than the preset single-trigger score threshold, the frequency modulation parameter value optimization process is directly started; this logic is suitable for scenarios where the unit's single frequency modulation performance is severely degraded, and can realize emergency response for parameter optimization and timely correct the problem of a sharp decline in frequency modulation performance.
[0023] The second type is the periodic average trigger judgment logic: calculate the historical frequency regulation effect score corresponding to all frequency regulation actions completed within a preset historical time period, and the average score of the current frequency regulation effect score. When the average score is lower than the preset periodic trigger score threshold, for example, the monthly average score is lower than the monthly average score threshold, the frequency regulation parameter value optimization process is initiated. This logic is suitable for scenarios where the frequency regulation performance continues to decline due to the slow drift of the unit's operating characteristics. It can avoid the false triggering of the optimization process caused by a single accidental fluctuation in operating conditions and ensure the stability of the unit's operation.
[0024] The preset score thresholds corresponding to the two judgment logics mentioned above can be set to different values independently to adapt to two different application scenarios, namely emergency optimization and regular iteration, while taking into account both the timeliness of parameter optimization and the stability of unit operation.
[0025] 104. If it is determined that activation is required, the pre-trained parameter optimization model is updated and trained based on the frequency modulation operating condition data, frequency modulation action data, and frequency modulation effect score to obtain the updated parameter optimization model.
[0026] 105. Based on the updated parameter value optimization model, generate a target frequency modulation parameter value combination for subsequent frequency modulation actions.
[0027] After determining in step 103 that the parameter value optimization process needs to be initiated, the process proceeds to step 104. The frequency modulation operating condition data, frequency modulation action data, and frequency modulation effect score obtained in this instance are used as new experience samples to incrementally update and train the pre-built parameter value optimization model, resulting in an updated parameter value optimization model. Subsequently, in step 105, a target frequency modulation parameter value combination for subsequent frequency modulation actions is generated based on the updated model.
[0028] The reason for updating and training the model is to allow it to continuously learn the actual frequency regulation response characteristics and parameter adaptation rules of the unit under different operating conditions, so that the model can dynamically evolve as the operating state of the unit changes, avoiding the decline in optimization ability due to model aging.
[0029] Based on the above Figure 1As can be seen from the implementation, the frequency regulation parameter value optimization method for coal-fired power units provided by this invention, in response to a frequency regulation action completed by the coal-fired power unit based on a combination of frequency regulation parameter values, acquires the frequency regulation operating condition data and frequency regulation action data. The frequency regulation action data includes the combination of frequency regulation parameter values between various frequency regulation parameters and the actual response data generated by the coal-fired power unit during the frequency regulation action based on the combination of frequency regulation parameter values. Subsequently, based on the frequency regulation operating condition data and the frequency regulation action data, the frequency regulation effect score of this frequency regulation action is determined, realizing a quantitative evaluation of the unit's single frequency regulation performance, thereby intuitively judging the degree of fit between the current combination of frequency regulation parameter values and the actual operating state of the unit. Furthermore, based on the quantified effect score, it is determined whether the frequency regulation parameter value optimization process needs to be initiated, so that the optimization action can be automatically triggered according to the actual performance, avoiding the lag and subjectivity of parameter adjustment. Once optimization is initiated, the pre-trained parameter optimization model is updated and trained using the frequency regulation operating condition data, frequency regulation action data, and frequency regulation effect score. This dynamic update mechanism overcomes the technical deficiency of fixed frequency regulation parameter values, which cannot keep up with changes in unit characteristics, leading to a decline in frequency regulation performance. Finally, the updated parameter optimization model generates a target combination of frequency regulation parameter values for subsequent frequency regulation actions. The generated parameter values better adapt to the unit's current actual operating state, which helps improve the adjustment accuracy and response performance of the unit's primary frequency regulation, thereby more effectively smoothing out grid frequency fluctuations and ensuring the safe and stable operation of the grid.
[0030] Furthermore, as a response to Figure 1 Further refinement and extension of the illustrated embodiment, this invention also provides another method for optimizing the frequency regulation parameter values of coal-fired power units, such as... Figure 2 As shown, the specific steps are as follows: 201. In response to the coal-fired unit completing a frequency regulation action based on the combination of frequency regulation parameter values, acquire the frequency regulation operating condition data and frequency regulation action data during the first frequency regulation action.
[0031] The implementation method of step 201 is the same as that of step 101, and can achieve the same technical effect and solve the same technical problem, so it will not be repeated here.
[0032] 202. Based on the frequency regulation operating condition data and frequency regulation action data, determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions.
[0033] In this embodiment, the frequency modulation speed score in the frequency modulation effect score can be obtained based on the actual load response curve, the adjustment time obtained from the actual load response curve, and the actual equivalent speed unequal rate. Specifically: The actual response data of the unit can be verified based on the target total change of the frequency regulation action (i.e., the maximum load adjustment that should be completed in one frequency regulation action). Specific verification items include: Determine whether the load change reflected in the actual load response curve within the first preset time period after the start of frequency regulation reaches the first preset proportion of the target total change. Determine whether the load change reaches the second preset ratio within the second preset time period after the start of frequency regulation; After the frequency regulation operation stabilizes, determine whether the actual equivalent speed unequal rate calculated based on the actual steady-state load contribution value and the frequency difference is within a preset compliance range.
[0034] If any of the above checks fails, the frequency modulation action is deemed not to meet the basic requirements for speed and accuracy, and the frequency modulation speed score is directly assigned to zero.
[0035] If all compliance checks pass, the system will further calculate the final score based on the adjustment time obtained from the actual load response curve. In this calculation, the shorter the adjustment time, the faster the unit responds, and the higher the frequency regulation speed score is obtained. Specifically, the score can be determined by setting specific value standards, that is, assigning a score corresponding to the range of adjustment time.
[0036] The frequency regulation speed score assesses the speed at which coal-fired power units respond to changes in grid frequency and the agility of their power regulation. The first preset duration / second preset duration refers to pre-set specific time points used to periodically examine the unit's response rate in the initial stage of frequency regulation.
[0037] The actual load response curve refers to the continuous trajectory data of the dynamic change of the active power of a coal-fired unit over time during the entire process of a frequency regulation operation. The load change refers to the difference between the actual load value of the unit at a specific time point during a frequency regulation operation and the reference load value at the start of the frequency regulation.
[0038] The first preset ratio / second preset ratio refers to the target percentage of the total change in the actual load of the unit that must be achieved within the corresponding preset time period.
[0039] The actual steady-state load contribution value refers to the increase or decrease in the actual active power output of the unit after a frequency regulation operation has entered a steady state compared to before the frequency regulation operation began. The adjustment time refers to the total time consumed during a frequency regulation operation, from the start of the change in the unit's active power until it finally reaches and stabilizes within the preset accuracy range of the target total change.
[0040] It should be noted that the frequency difference mentioned in this invention is the difference between the group frequency and 50Hz at the start of the first frequency modulation.
[0041] Meanwhile, the frequency modulation accuracy score can be determined based on the load deviation between the actual steady-state load value and the theoretical target load value after a single frequency modulation operation. The theoretical target load value is calculated based on the frequency difference and the speed unequal rate setting value. The smaller the load deviation, the higher the frequency modulation accuracy score. Alternatively, a graded matching mode of error range and score can be adopted, that is, different load deviation ranges are preset and correspond to different score levels, with a higher score level corresponding to a smaller deviation range.
[0042] The actual steady-state load value is a single value representing the final static result extracted from the stable section at the end of the actual load response curve. It is usually obtained by averaging or fitting a series of data points after the curve enters the convergence state.
[0043] Meanwhile, the frequency regulation safety score can be determined based on the maximum fluctuation amplitude of the main steam pressure after frequency regulation stabilization and the pressure fluctuation recovery time. A pressure over-limit rejection mechanism can be set: if the maximum fluctuation amplitude exceeds a preset safety threshold (e.g., 0.3 MPa), the frequency regulation safety score will be directly determined as the lowest score; conversely, when the pressure fluctuation does not exceed 0.3 MPa after frequency regulation stabilization, the score will not be determined as the lowest score; a graded matching mode between numerical values and scores can also be adopted, that is, different fluctuation amplitude and recovery time ranges are preset, corresponding to different score levels. The smaller the fluctuation amplitude and the shorter the recovery time, the higher the corresponding score level.
[0044] The maximum fluctuation amplitude is the maximum absolute value of the difference between each actual main steam pressure monitoring value and the reference main steam pressure after the frequency regulation stabilizes; the pressure fluctuation recovery time is the time difference from the time the frequency regulation action stabilizes until the real-time monitoring value of the main steam pressure remains stable within the preset allowable deviation range of the reference main steam pressure.
[0045] Simultaneously, the overall valve position change amplitude at the start of frequency modulation can be extracted based on the actual overall valve position opening curve (for example, the economy is best when the opening is between 75% and 85%, the regulating ability is basically lost when it exceeds 85%, and the economy is poor when it is less than 75%), and the frequency modulation economy score can be determined accordingly. The overall valve position change amplitude characterizes the severity of the frequency modulation action; the smaller the change amplitude, the smoother the frequency modulation action, the lower the mechanical loss of the equipment, and the higher the corresponding frequency modulation economy score. This step can also adopt a graded matching mode of numerical values and scores, that is, preset different valve position change amplitude ranges and corresponding score levels.
[0046] It is important to note that all scores need to be normalized. This involves mapping scores of different dimensions and orders of magnitude (such as accuracy, stability, safety, and economy) to the same numerical range (e.g., [0,1] or [0,100]) through mathematical transformation. This eliminates biases caused by differences in units or values between indicators, ensuring the objectivity of the comprehensive evaluation function calculation results. Finally, the weighted average score among the FM speed, accuracy, safety, and economy scores is determined as the FM performance score.
[0047] The actual integrated valve position opening curve records the continuous data trajectory of the equivalent valve position opening of the integrated steam inlet of all main control valves of the steam turbine as a function of time. Based on this curve, the amplitude of the integrated valve position change during the frequency regulation process can be calculated. This amplitude is a quantitative indicator used to measure the total action stroke or flow curve inflection point of the control valve when responding to the command. When the integrated valve position is at the inflection point, the unit control valve is prone to frequent vibration, which can cause low-frequency oscillations and lead to serious grid accidents.
[0048] 203. Based on the frequency modulation effect score, determine whether it is necessary to start the frequency modulation parameter value optimization process.
[0049] In this embodiment, it is preferable to obtain the historical frequency regulation effect scores obtained when the coal-fired power unit performs one frequency regulation action within a preset historical time period; then, the average effect score between the historical frequency regulation effect scores and the average effect score is determined. If the average effect score is less than a preset score threshold, the frequency regulation parameter value optimization process is initiated; if the average effect score is not less than the preset score threshold, the frequency regulation parameter value optimization process is not initiated.
[0050] In this way, when the average performance score is lower than the threshold, it indicates that the current frequency regulation performance of the unit can no longer meet the grid requirements or there is room for optimization. At this time, the parameter optimization process is automatically started, which can correct the frequency regulation strategy in a timely manner and improve the frequency regulation response capability of the unit. When the average performance score reaches or exceeds the threshold, it means that the current parameter settings can meet the frequency regulation requirements and there is no need for frequent adjustments, thereby avoiding unnecessary parameter disturbances and ensuring the stability of unit operation.
[0051] 204. If it is determined that activation is required, the pre-trained parameter optimization model is updated and trained based on the frequency modulation operating condition data, frequency modulation action data, and frequency modulation effect score to obtain the updated parameter optimization model.
[0052] 205. Based on the updated parameter value optimization model, generate a target frequency modulation parameter value combination for subsequent frequency modulation actions.
[0053] In step 204, the frequency modulation effect score can be compared with a preset sample score threshold. If the frequency modulation effect score is not less than the preset sample score threshold, the frequency modulation operating condition data, frequency modulation action data, and frequency modulation effect score are marked as positive samples. If the frequency modulation effect score is less than the preset sample judgment threshold, the frequency modulation operating condition data, frequency modulation action data, and frequency modulation effect score are marked as negative samples.
[0054] This allows the training dataset to be updated based on positive or negative samples, and the updated training dataset can be used to train the parameter optimization model to obtain the updated parameter optimization model. The weights of the newly added samples in the training dataset are configured to be the weights of the original samples in the large training dataset.
[0055] Subsequently, in step 205, the predicted target frequency regulation condition data for the subsequent frequency regulation operation of the coal-fired unit can be determined. The predicted target frequency regulation condition data includes different load segments and different frequency regulation directions. The predicted target frequency modulation (FM) conditions are input into the updated parameter optimization model. The updated model, with the optimization objective of maximizing the FM performance score, outputs a combination of target FM parameter values corresponding to each predicted FM condition. This mechanism uses positive and negative sample labels to distinguish between successful experiences and lessons learned in the frequency modulation process, and assigns higher training weights to newly added samples, so that the parameter optimization model can quickly adapt to the current operating characteristics and aging of the unit, and prioritize the absorption of the latest operating rules, thereby significantly improving the model's fitting accuracy and generalization ability to the current operating conditions. On the other hand, by pre-setting the predicted target frequency regulation operating condition data covering different load segments and frequency regulation directions, and driving the updated model to perform optimization calculations with the goal of maximizing the frequency regulation effect score, it is possible to generate and store the optimal combination of frequency regulation parameter values for different operating scenarios in advance. This ensures that when the unit faces complex and ever-changing grid frequency fluctuations in the future, it can immediately call the optimal strategy, ensuring the speed and accuracy of frequency regulation response while taking into account the safety and economy of the equipment.
[0056] 206. Send the predicted target frequency modulation operating condition data and the target frequency modulation parameter value combination to the human-machine interface to confirm with the staff whether to use the target frequency modulation parameter value combination as the frequency modulation parameter value combination to perform frequency modulation action when encountering each predicted target frequency modulation operating condition data in the future.
[0057] Since the target frequency regulation parameter value combination output by the updated parameter value optimization model is based on the theoretical optimal solution generated by historical data and algorithm model, although this combination can theoretically maximize the frequency regulation effect score, directly applying parameters without human review may lead to the risk of unstable unit operation. Therefore, it is necessary to send each predicted target frequency regulation operating condition data and the corresponding target frequency regulation parameter value combination to the human-machine interface, and introduce the professional experience of the staff to conduct the final safety and feasibility confirmation. After receiving the manual confirmation instruction, the target frequency regulation parameter value combination will be configured as the execution parameter for performing frequency regulation action when the corresponding predicted target frequency regulation operating condition data is encountered in the future, thereby improving frequency regulation performance while ensuring the safety of unit operation.
[0058] However, if a manual rejection instruction is received, the rejected parameter value combination in the target frequency modulation parameter value combination targeted by the manual rejection instruction is obtained; the training dataset and training constraints of the parameter value optimization model are updated according to the rejected parameter value combination.
[0059] If a manual rejection instruction is received, it indicates that although the parameters generated by the model score highly in theoretical calculations, they may touch upon potential safety hazards or violate the on-site operational experience of operators. Therefore, these rejected parameter combinations must be extracted and re-input into the training set as "negative samples" or "hard constraints" to update the model. In this way, the model is forced to "memorize" and avoid these manually rejected erroneous paths, transforming the implicit experience of operators into explicit constraints for the model. This prevents the model from repeatedly recommending such parameters in the future, ensuring that the optimization results not only have high theoretical scores but are also safe and feasible in practical applications.
[0060] This negative sample training mechanism essentially labels negative samples as constraint samples with strong penalty labels and introduces a corresponding penalty term into the model's loss function to reduce the spatial weight of constraint samples and their neighborhoods in the parameter generation probability distribution during model training. Alternatively, a logical gating filter can be set at the model's output. When the recommended parameters generated by the model fall into the historical feature range of the constraint samples, a backoff mechanism is triggered to block the recommended parameters. In contrast, the loss function weight of positive samples is increased during model training to enhance their spatial weight in the parameter generation probability distribution and guide the model to converge to the parameter region that has been verified in practice.
[0061] Furthermore, as a response to the above Figure 1 In addition to the implementation of the method shown, this embodiment of the invention also provides a frequency regulation parameter value optimization device for coal-fired power units, used for optimizing the above-mentioned frequency regulation parameters. Figure 1The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 3 As shown, the device includes: The acquisition unit 301 is used to acquire frequency regulation condition data and frequency regulation action data during a frequency regulation action in response to the coal-fired power unit completing a frequency regulation action based on the combination of frequency regulation parameter values. The frequency regulation action data includes the combination of frequency regulation parameter values between each frequency regulation parameter and the actual response data generated by the coal-fired power unit during the frequency regulation action based on the combination of frequency regulation parameter values. The scoring determination unit 302 is used to determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions based on the frequency regulation condition data and frequency regulation action data acquired by the acquisition unit 301. The optimization judgment unit 303 is used to determine whether the frequency modulation parameter value optimization process needs to be started based on the frequency modulation effect score determined by the score determination unit 302. The model training unit 304 is used to update and train the pre-trained parameter value optimization model based on the frequency modulation operating condition data, the frequency modulation action data and the frequency modulation effect score if the judgment result of the optimization judgment unit 303 determines that it needs to be started, so as to obtain the updated parameter value optimization model. The numerical optimization unit 305 is used to optimize the model based on the updated parameter values obtained by the model training unit 304, and generate a target frequency modulation parameter value combination for subsequent frequency modulation actions.
[0062] Furthermore, as a response to the above Figure 2 In addition to the method shown, this embodiment of the invention also provides another frequency regulation parameter value optimization device for coal-fired power units, used for optimizing the above-mentioned frequency regulation parameters. Figure 2 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 4 As shown, the device includes: The acquisition unit 301 is used to acquire frequency regulation condition data and frequency regulation action data during a frequency regulation action in response to the coal-fired power unit completing a frequency regulation action based on the combination of frequency regulation parameter values. The frequency regulation action data includes the combination of frequency regulation parameter values between each frequency regulation parameter and the actual response data generated by the coal-fired power unit during the frequency regulation action based on the combination of frequency regulation parameter values. The scoring determination unit 302 is used to determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions based on the frequency regulation condition data and frequency regulation action data acquired by the acquisition unit 301. The optimization judgment unit 303 is used to determine whether the frequency modulation parameter value optimization process needs to be started based on the frequency modulation effect score determined by the score determination unit 302. The model training unit 304 is used to update and train the pre-trained parameter value optimization model based on the frequency modulation operating condition data, the frequency modulation action data and the frequency modulation effect score if the judgment result of the optimization judgment unit 303 determines that it needs to be started, so as to obtain the updated parameter value optimization model. The numerical optimization unit 305 is used to optimize the model based on the updated parameter values obtained by the model training unit 304, and generate a target frequency modulation parameter value combination for subsequent frequency modulation actions.
[0063] In one optional implementation, the optimization judgment unit 303 is specifically used for: Obtain the historical frequency regulation effect scores obtained when a coal-fired power unit performs one frequency regulation action within a preset historical time period; Determine the historical frequency modulation effect score and the average effect score among the frequency modulation effect scores; If the average effect score is less than the preset score threshold, then the frequency tuning parameter value optimization process is initiated. If the average effect score is not less than the preset score threshold, then the frequency modulation parameter value optimization process will not be started.
[0064] In one optional implementation, the model training unit 304 is specifically used for: The frequency modulation effect score is compared with a preset sample score threshold. If the frequency modulation effect score is not less than the preset sample score threshold, then the frequency modulation condition data, the frequency modulation action data, and the frequency modulation effect score are marked as positive samples. If the frequency modulation effect score is less than the preset sample judgment threshold, then the frequency modulation condition data, the frequency modulation action data, and the frequency modulation effect score are marked as negative samples. The training dataset is updated based on the positive or negative samples, and the parameter value optimization model is trained using the updated training dataset to obtain the updated parameter value optimization model. The weights of the newly added samples in the training dataset are configured to be the weights of the original samples in the large training dataset.
[0065] In one optional implementation, the numerical optimization unit 305 is specifically used for: Determine the predicted target frequency regulation conditions for each subsequent frequency regulation operation of the coal-fired power unit, wherein the predicted target frequency regulation conditions include different load segments and different frequency regulation directions; The predicted target frequency modulation operating conditions data are input into the updated parameter value optimization model. The updated parameter value optimization model takes maximizing the frequency modulation effect score as the optimization objective and outputs the target frequency modulation parameter value combination corresponding to each predicted target frequency modulation operating condition.
[0066] In an optional implementation, after the numerical optimization unit 305 obtains the target frequency modulation parameter value combination corresponding to each predicted target frequency modulation condition data, the device further includes a parameter value confirmation unit 306, which is specifically used for: Send the predicted target frequency modulation operating condition data and the target frequency modulation parameter value combination to the human-machine interface to confirm with the staff whether to use the target frequency modulation parameter value combination as the frequency modulation parameter value combination to perform frequency modulation action when encountering each predicted target frequency modulation operating condition data in the future. In response to the received manual confirmation instruction, the target frequency modulation parameter value combination is configured as the execution parameter for performing frequency modulation action when the corresponding predicted target frequency modulation operating condition data is encountered in the future.
[0067] In one optional implementation, if the parameter value confirmation unit 306 receives a manual rejection instruction, it is specifically used for: Obtain the rejected parameter value combination from the target frequency modulation parameter value combination targeted by the manual rejection instruction; The training dataset and training constraints of the parameter value optimization model are updated based on the rejected parameter value combinations.
[0068] In one optional implementation, the score determination unit 302 is specifically used for: Obtain the reference unit load, reference main steam pressure, initial integrated valve position opening, and frequency difference at the start of frequency regulation from the frequency regulation operating condition data; Obtain the speed unequal rate setting value, frequency modulation direction, and target frequency modulation load command from each frequency modulation parameter value in the frequency modulation action data; The actual load response curve, the adjustment time, the actual main steam pressure curve, and the actual comprehensive valve position opening curve obtained from the actual load response curve of the frequency regulation action data are acquired. The frequency regulation speed score, frequency regulation accuracy score, frequency regulation safety score, and frequency regulation economy score are determined based on the benchmark unit load, benchmark main steam pressure, initial comprehensive valve position opening, frequency difference, speed inequality setpoint, actual load response curve, adjustment time obtained from the actual load response curve, actual main steam pressure curve, and actual comprehensive valve position opening curve. The weighted average score among the frequency modulation speed score, frequency modulation accuracy score, frequency modulation security score, and frequency modulation economy score is determined as the frequency modulation effect score.
[0069] Furthermore, embodiments of the present invention also provide a storage medium for storing a computer program, wherein the computer program, when running, controls the device where the storage medium is located to execute the above-described... Figure 1-2 The method for optimizing the frequency regulation parameters of coal-fired power units described herein.
[0070] Furthermore, embodiments of the present invention also provide a processor for running a program, wherein the program executes the above-described... Figure 1-2 The method for optimizing the frequency regulation parameters of coal-fired power units described herein.
[0071] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.
[0072] It is understood that the relevant features in the above methods and apparatus can be referenced interchangeably. Furthermore, the terms "first," "second," etc., in the above embodiments are used to distinguish between embodiments and do not represent the superiority or inferiority of any particular embodiment.
[0073] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0074] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages is for the purpose of disclosing the best mode of implementation of the invention.
[0075] In addition, the memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.
[0076] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0077] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0078] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0079] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0080] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0081] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.
[0082] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.
[0083] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0084] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0085] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A method for optimizing frequency modulation parameter values of a coal-fired unit, characterized in that, The method includes: In response to the coal-fired unit completing a frequency regulation action based on the combination of frequency regulation parameter values, acquire the frequency regulation condition data and frequency regulation action data during the first frequency regulation action; Based on the frequency regulation operating condition data and the frequency regulation action data, determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions; Based on the frequency modulation effect score, determine whether it is necessary to initiate the frequency modulation parameter value optimization process; If it is determined that activation is required, the pre-trained parameter value optimization model is updated and trained based on the frequency modulation operating condition data, the frequency modulation action data, and the frequency modulation effect score to obtain the updated parameter value optimization model. Based on the updated parameter value optimization model, a target frequency modulation parameter value combination is generated for subsequent frequency modulation actions.
2. The method of claim 1, wherein, Based on the frequency modulation effect score, determine whether the frequency modulation parameter value optimization process needs to be initiated, including: Obtain the historical frequency regulation effect scores obtained when a coal-fired power unit performs one frequency regulation action within a preset historical time period; Determine the historical frequency modulation effect score and the average effect score among the frequency modulation effect scores; If the average effect score is less than the preset score threshold, then the frequency tuning parameter value optimization process is initiated. If the average effect score is not less than the preset score threshold, then the frequency modulation parameter value optimization process will not be started.
3. The method of claim 1, wherein, Based on the frequency modulation operating condition data, the frequency modulation action data, and the frequency modulation effect score, the pre-trained parameter value optimization model is updated and trained to obtain the updated parameter value optimization model, including: The frequency modulation effect score is compared with a preset sample score threshold. If the frequency modulation effect score is not less than the preset sample score threshold, then the frequency modulation condition data, the frequency modulation action data, and the frequency modulation effect score are marked as positive samples. If the frequency modulation effect score is less than the preset sample judgment threshold, then the frequency modulation condition data, the frequency modulation action data, and the frequency modulation effect score are marked as negative samples. The training dataset is updated based on the positive or negative samples, and the parameter value optimization model is trained using the updated training dataset to obtain the updated parameter value optimization model. The weights of the newly added samples in the training dataset are configured to be the weights of the original samples in the large training dataset.
4. The method of claim 1, wherein, Based on the updated parameter value optimization model, a target frequency modulation parameter value combination for subsequent frequency modulation actions is generated, including: Determine the predicted target frequency regulation conditions for each subsequent frequency regulation operation of the coal-fired power unit, wherein the predicted target frequency regulation conditions include different load segments and different frequency regulation directions; The predicted target frequency modulation operating conditions data are input into the updated parameter value optimization model. The updated parameter value optimization model takes maximizing the frequency modulation effect score as the optimization objective and outputs the target frequency modulation parameter value combination corresponding to each predicted target frequency modulation operating condition.
5. The method according to any one of claims 1-4, characterized in that, After obtaining the combination of target frequency modulation parameter values corresponding to each predicted target frequency modulation condition data, the method further includes: Send the predicted target frequency modulation operating condition data and the target frequency modulation parameter value combination to the human-machine interface to confirm with the staff whether to use the target frequency modulation parameter value combination as the frequency modulation parameter value combination to perform frequency modulation action when encountering each predicted target frequency modulation operating condition data in the future. In response to the received manual confirmation instruction, the target frequency modulation parameter value combination is configured as the execution parameter for performing frequency modulation action when the corresponding predicted target frequency modulation operating condition data is encountered in the future.
6. The method of claim 5, wherein, If a manual rejection instruction is received, the method further includes: Obtain the rejected parameter value combination from the target frequency modulation parameter value combination targeted by the manual rejection instruction; The training dataset and training constraints of the parameter value optimization model are updated based on the rejected parameter value combinations.
7. The method according to any one of claims 1-4, characterized in that, Based on the frequency regulation operating condition data and the frequency regulation action data, determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions, including: Obtain the reference unit load, reference main steam pressure, initial integrated valve position opening, and frequency difference at the start of frequency regulation from the frequency regulation operating condition data; Obtain the speed unequal rate setting value, frequency modulation direction, and target frequency modulation load command from each frequency modulation parameter value in the frequency modulation action data; The actual load response curve, the adjustment time, the actual main steam pressure curve, and the actual comprehensive valve position opening curve obtained from the actual load response curve of the frequency regulation action data are acquired. The frequency regulation speed score, frequency regulation accuracy score, frequency regulation safety score, and frequency regulation economy score are determined based on the benchmark unit load, benchmark main steam pressure, initial comprehensive valve position opening, frequency difference, speed inequality setpoint, actual load response curve, adjustment time obtained from the actual load response curve, actual main steam pressure curve, and actual comprehensive valve position opening curve. The weighted average score among the frequency modulation speed score, frequency modulation accuracy score, frequency modulation security score, and frequency modulation economy score is determined as the frequency modulation effect score.
8. A device for optimizing frequency modulation parameter values of a coal-fired unit, characterized in that, The device includes: The acquisition unit is used to respond to the coal-fired power unit completing a frequency regulation action based on the combination of frequency regulation parameter values, and to acquire frequency regulation condition data and frequency regulation action data during the frequency regulation action. The frequency regulation action data includes the combination of frequency regulation parameter values between various frequency regulation parameters and the actual response data generated by the coal-fired power unit based on the combination of frequency regulation parameter values during the frequency regulation action. The scoring determination unit is used to determine the frequency regulation effect score of the coal-fired unit when performing frequency regulation actions based on the frequency regulation condition data and frequency regulation action data obtained by the acquisition unit. An optimization judgment unit is used to determine whether the frequency modulation effect score determined by the score determination unit needs to be started, and whether the frequency modulation parameter value optimization process needs to be initiated. The model training unit is used to update and train the pre-trained parameter value optimization model based on the frequency modulation operating condition data, the frequency modulation action data, and the frequency modulation effect score if the optimization judgment unit determines that it needs to be started, so as to obtain the updated parameter value optimization model. The numerical optimization unit is used to optimize the model based on the updated parameter values obtained by the model training unit, and generate a combination of target frequency modulation parameter values for subsequent frequency modulation actions.
9. A storage medium, characterized by The storage medium comprises a stored program, wherein the program controls a device where the storage medium is located to execute the method for optimizing frequency modulation parameter values of a coal-fired generating unit according to any one of claims 1 to 7 when the program is running.
10. A processor, comprising: The processor is configured to run a program, wherein the program executes the method for optimizing frequency modulation parameter values of a coal-fired generating unit according to any one of claims 1 to 7 when the program is running.
