Power grid peak and frequency regulation adaptive optimization platform

Abstract

The present invention discloses an adaptive optimization platform for power grid peak-shaving and frequency regulation, which belongs to the field of power grid peak-shaving and frequency regulation, and includes: a peak gap prediction module for determining multiple peak gap time zones and multiple peak power generation gaps; a peak-shaving and frequency regulation decision module for generating multiple peak-shaving and frequency regulation strategies; a peak-shaving and frequency regulation loss evaluation module for obtaining multiple peak-shaving and frequency regulation loss coefficients; a peak-shaving and frequency regulation loss judgment module for generating multiple peak-shaving and frequency regulation loss judgment results; a peak-shaving and frequency regulation optimization module for generating a peak-shaving and frequency regulation optimization strategy set; and a peak-shaving and frequency regulation execution module for performing peak-shaving and frequency regulation on the power grid. The present invention solves the technical problem in the prior art that power grid peak-shaving and frequency regulation is difficult to effectively deal with power grid peak generation gaps, resulting in unstable power grid operation, and achieves the technical effect of effectively dealing with power grid peak gaps and ensuring stable power grid operation through adaptive optimization of power grid peak-shaving and frequency regulation.

Description

Technical Field

[0001] The present invention relates to the field of power grid peak-shaving and frequency regulation, and in particular to a power grid peak-shaving and frequency regulation adaptive optimization platform. Background Art

[0002] Safe and stable grid operation is a core requirement of the power system, and grid peak and frequency regulation plays a crucial role. However, existing technologies often fail to effectively address peak power generation shortfalls, leading to numerous operational challenges for the grid. Peak power generation shortfalls occur when grid load increases dramatically but power generation is insufficient to meet demand. If peak and frequency regulation are not implemented promptly and effectively, grid frequency and voltage may fluctuate significantly, potentially causing widespread power outages and significant losses. Traditional grid peak and frequency regulation relies primarily on the experience of grid dispatchers and follows pre-set strategies. This approach struggles to quickly respond to real-time changes in grid peak conditions, resulting in low accuracy and efficiency. Furthermore, different peak power generation shortfalls often require different peak and frequency regulation strategies, and existing fixed strategies lack the flexibility to address them. Consequently, existing technologies suffer from the technical problem of grid peak and frequency regulation failing to effectively address peak power generation shortfalls, leading to unstable grid operation. Summary of the Invention

[0003] This application provides a grid peak-shaving and frequency-regulating adaptive optimization platform, aiming to solve the technical problem in the existing technology that grid peak-shaving and frequency-regulation cannot effectively cope with the peak power generation gap of the grid, resulting in unstable grid operation.

[0004] The power grid peak-shaving and frequency-regulation adaptive optimization platform disclosed in the present application includes: a peak gap prediction module for predicting the peak gap of the first power generation side of the power grid according to the future time zone window, and determining multiple peak gap time zones and multiple peak power generation gaps, wherein the first power generation side is in a grid-connected operation state and the second power generation side of the power grid is in an island operation state; a peak-shaving and frequency-regulation decision module for making peak-shaving and frequency-regulation decisions on the power grid according to the second power generation side based on the multiple peak gap time zones and multiple peak power generation gaps, and generating multiple peak-shaving and frequency-regulation strategies; a peak-shaving and frequency-regulation loss evaluation module, Used to predict and evaluate the losses of multiple peak-shaving and frequency regulation strategies, and obtain multiple peak-shaving and frequency regulation loss coefficients; the peak-shaving and frequency regulation loss judgment module is used to judge whether multiple peak-shaving and frequency regulation loss coefficients are less than the peak-shaving and frequency regulation loss threshold, and generate multiple peak-shaving and frequency regulation loss judgment results; the peak-shaving and frequency regulation optimization module is used to optimize and adjust multiple peak-shaving and frequency regulation strategies based on multiple peak-shaving and frequency regulation loss judgment results, and generate a peak-shaving and frequency regulation optimization strategy set; the peak-shaving and frequency regulation execution module is used to perform peak-shaving and frequency regulation on the power grid based on the second power generation side and the peak-shaving and frequency regulation optimization strategy set based on multiple peak gap time zones.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] The peak gap prediction module predicts peak gaps for the first generation side of the grid in grid-connected operation based on future time zone windows, determines multiple peak gap time zones and corresponding peak generation gaps, and provides a basis for subsequent peak-shaving and frequency regulation decisions. The peak-shaving and frequency regulation decision module makes peak-shaving and frequency regulation decisions based on the predicted multiple peak gap time zones and peak generation gaps, while also considering the operating status of the second generation side of the grid in islanded operation. Multiple alternative peak-shaving and frequency regulation strategies are generated. Targeted peak-shaving and frequency regulation strategies are formulated based on different peak gap situations, improving the flexibility of grid peak-shaving and frequency regulation. The peak-shaving and frequency regulation loss evaluation module predicts and evaluates the losses of multiple alternative peak-shaving and frequency regulation strategies, obtains the peak-shaving and frequency regulation loss coefficient corresponding to each strategy, and quantitatively evaluates the economy and rationality of each peak-shaving and frequency regulation strategy, providing a reference for strategy optimization. The peak-shaving and frequency regulation loss judgment module compares the peak-shaving and frequency regulation loss coefficient of each strategy with a preset loss threshold to determine whether it meets the requirements. The corresponding peak-shaving and frequency regulation loss judgment result is output, and the strategy with lower loss is selected. Through the peak-shaving and frequency regulation optimization module, based on the peak-shaving and frequency regulation loss judgment results, multiple alternative peak-shaving and frequency regulation strategies are optimized and combined to generate the peak-shaving and frequency regulation optimization strategy set with the best comprehensive performance. Under the premise of taking into account the safe and stable operation of the power grid, the peak-shaving and frequency regulation strategy combination with the greatest overall economic benefit is selected; when the predicted peak gap is about to come, the peak-shaving and frequency regulation execution module controls the second power generation side to perform real-time peak-shaving and frequency regulation on the power grid based on the optimized peak-shaving and frequency regulation optimization strategy set, dynamically matches the demand changes of the power grid, and effectively alleviates the peak power generation gap of the power grid. This technical solution solves the technical problem that the power grid peak-shaving and frequency regulation in the existing technology is difficult to effectively cope with the peak power generation gap of the power grid, resulting in unstable power grid operation, and achieves the technical effect of effectively coping with the peak gap of the power grid and ensuring the stable operation of the power grid through adaptive optimization of the power grid peak-shaving and frequency regulation.

[0007] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A schematic structural diagram of a power grid peak and frequency regulation adaptive optimization platform is provided for the embodiment of the present application;

[0009] Figure 2 A flowchart of generating multiple peak-shaving and frequency-regulating strategies in a power grid peak-shaving and frequency-regulating adaptive optimization platform is provided for an embodiment of the present application.

[0010] Explanation of the accompanying symbols: peak gap prediction module 11, peak-shaving and frequency regulation decision module 12, peak-shaving and frequency regulation loss evaluation module 13, peak-shaving and frequency regulation loss judgment module 14, peak-shaving and frequency regulation optimization module 15, peak-shaving and frequency regulation execution module 16. DETAILED DESCRIPTION

[0011] The overall idea of ​​the technical solution provided by this application is as follows:

[0012] The embodiment of the present application provides an adaptive optimization platform for peak-shaving and frequency regulation of power grids. First, the peak gap prediction module is used to predict the peak power generation gap that may occur in the power grid within the future time zone window, and to identify and quantify the power supply and demand imbalance risk faced by the first power generation side of the power grid in advance. Then, the peak-shaving and frequency regulation decision module automatically generates multiple alternative peak-shaving and frequency regulation strategies based on the prediction results and the actual operating status of the second power generation side of the power grid, and evaluates each strategy through the peak-shaving and frequency regulation loss evaluation module, and uses the peak-shaving and frequency regulation loss judgment module to screen out strategies that meet the loss threshold requirements. Then, the peak-shaving and frequency regulation optimization module searches and determines the peak-shaving and frequency regulation optimization strategy set with the best comprehensive performance under the premise of meeting the constraints of safe and stable operation of the power grid. Afterwards, when the predicted peak gap is about to arrive, the peak-shaving and frequency regulation execution module controls the second power generation side of the power grid to perform real-time peak-shaving and frequency regulation on the power grid based on the peak-shaving and frequency regulation optimization strategy set, dynamically matching the demand changes of the power grid, thereby effectively alleviating the peak power generation gap problem on the first power generation side of the power grid.

[0013] After introducing the basic principles of the present application, the following will specifically introduce the non-limiting implementation methods of the present application in conjunction with the drawings in the specification.

[0014] like Figure 1 As shown, the embodiment of the present application provides a power grid peak and frequency regulation adaptive optimization platform, which includes:

[0015] The peak gap prediction module 11 is used to predict the peak gap of the first power generation side of the power grid according to the future time zone window, and determine multiple peak gap time zones and multiple peak power generation gaps, wherein the first power generation side is in a grid-connected operation state and the second power generation side of the power grid is in an island operation state.

[0016] Specifically, the peak gap prediction module 11 is used to perform a peak gap prediction analysis on the first power generation side in the grid-connected operation state according to a time zone window in the future.

[0017] First, the peak-gap prediction module 11 obtains time zone window information within a certain future time range, which serves as the future time zone window. This future time zone window is pre-set based on the grid's operating characteristics and load variations, or it can be dynamically adjusted based on real-time data. Then, based on this future time zone window, the peak-gap prediction module 11 analyzes and compares the power generation capacity and load demand of the first generation side in grid-connected operation. By calculating the difference between the first generation side's generated power and load power, the module determines the time intervals within the future time zone window during which the grid may experience power supply gaps, i.e., multiple peak-gap time zones. Furthermore, the module determines the difference between the first generation side's generated power and load power within each peak-gap time zone, i.e., multiple peak-gap time zones. There are two generation sides in different operating states: a first generation side and a second generation side. The first generation side is grid-connected and its generated power can be directly fed into the grid. The second generation side is in islanded operation and its generated power cannot be directly fed into the grid, but it can provide auxiliary support for grid peak and frequency regulation.

[0018] Through the peak gap prediction module 11, the supply and demand imbalance of the power grid in the future period is predicted, laying the foundation for optimizing scheduling and scientific peak and frequency regulation, and improving the operation efficiency and power supply reliability of the power grid.

[0019] The peak-shaving and frequency regulation decision module 12 is used to make peak-shaving and frequency regulation decisions on the power grid according to the second power generation side based on the multiple peak gap time zones and the multiple peak power generation gaps, and generate multiple peak-shaving and frequency regulation strategies.

[0020] Specifically, the peak-shaving and frequency regulation decision module 12 first receives the multiple peak-shaving time zones and multiple peak-generation gaps output by the peak-shaving gap prediction module 11. This data reflects the power supply gaps that the power grid may face in different time intervals over the next period of time. Subsequently, the peak-shaving and frequency regulation decision module 12 comprehensively considers the actual operating status and peak-shaving and frequency regulation capabilities of the second generation side. Although the second generation side operates in an isolated state, its power generation resources can provide certain support for the power grid's peak-shaving and frequency regulation. The peak-shaving and frequency regulation decision module 12 analyzes parameters such as the second generation side's power generation capacity, response speed, and regulation accuracy to determine the extent to which the second generation side can participate in the power grid's peak-shaving and frequency regulation process. Based on the peak-shaving gap information and the status of the second generation side, the peak-shaving and frequency regulation decision module 12 begins to formulate a corresponding peak-shaving and frequency regulation strategy for each peak-shaving time zone, taking into account the size of the peak-generation gap. The peak-shaving and frequency regulation strategy includes the second generation side's output regulation plan, regulation time sequence, and regulation power allocation. Each peak-period gap time zone corresponds to an alternative peak-shaving and frequency regulation strategy, and the peak-shaving and frequency regulation decision module 12 generates multiple peak-shaving and frequency regulation strategies for subsequent evaluation and optimization.

[0021] Through the strategy generation of the peak-shaving and frequency regulation decision module 12, multiple peak-shaving and frequency regulation strategies are obtained, which lays the foundation for subsequent strategy optimization and execution, so that the power grid can quickly make corresponding peak-shaving and frequency regulation responses when facing peak period gaps, thereby ensuring the safe and stable operation of the power system.

[0022] The peak-shaving and frequency-regulation loss evaluation module 13 is configured to perform loss prediction evaluation on the multiple peak-shaving and frequency-regulation strategies to obtain multiple peak-shaving and frequency-regulation loss coefficients.

[0023] Specifically, after receiving multiple peak-shaving and frequency regulation strategies generated by the peak-shaving and frequency regulation decision module 12, the peak-shaving and frequency regulation loss assessment module 13 performs loss prediction and assessment for each peak-shaving and frequency regulation strategy. For example, the peak-shaving and frequency regulation loss assessment module 13 considers four aspects: power quality loss, power generation resource loss, equipment wear loss, and economic benefit loss. Regarding power quality loss, the peak-shaving and frequency regulation loss assessment module 13 analyzes changes in power quality indicators such as voltage and frequency of the power grid after the implementation of the peak-shaving and frequency regulation strategy, predicts possible power supply quality and stability issues, and provides corresponding loss assessment results. Regarding power generation resource loss, the peak-shaving and frequency regulation loss assessment module 13 analyzes the impact of the power output adjustment on the power generation resource utilization efficiency when the second power generation side implements the peak-shaving and frequency regulation strategy, and estimates the corresponding energy loss data. Regarding equipment wear loss, the peak-shaving and frequency regulation loss assessment module 13 assesses the accelerated effect of frequent output adjustment operations on the wear and aging of the power generation equipment on the second power generation side during the implementation of the peak-shaving and frequency regulation strategy, and predicts the impact on the equipment service life and the related maintenance and replacement costs. For the loss of economic benefits, the peak-shaving and frequency regulation loss evaluation module 13 analyzes the changes in operating costs and benefits of the power grid and the second power generation side after the implementation of the peak-shaving and frequency regulation strategy, and evaluates the economic impact of the strategy. After completing the loss assessment of the above four aspects, the peak-shaving and frequency regulation loss evaluation module 13 obtains four corresponding loss index data for each peak-shaving and frequency regulation strategy. Next, the peak-shaving and frequency regulation loss evaluation module 13 performs weighted processing on these four loss indicators. Specifically, different weight coefficients are given according to the degree of influence of each loss indicator on the peak-shaving and frequency regulation effect of the power grid, and then each loss indicator is multiplied by its weight coefficient, and the sum is obtained to obtain a comprehensive peak-shaving and frequency regulation loss coefficient, which is used to reflect the overall loss level of the peak-shaving and frequency regulation strategy.

[0024] Through the above process, the peak-shaving and frequency regulation loss evaluation module 13 associates each peak-shaving and frequency regulation strategy with a peak-shaving and frequency regulation loss coefficient, obtains multiple peak-shaving and frequency regulation loss coefficients, and forms a mapping relationship between the peak-shaving and frequency regulation strategy and its peak-shaving and frequency regulation loss coefficient, laying the foundation for subsequent peak-shaving and frequency regulation strategy screening and optimization.

[0025] The peak-shaving and frequency modulation loss judgment module 14 is configured to judge whether the plurality of peak-shaving and frequency modulation loss coefficients are less than a peak-shaving and frequency modulation loss threshold, and generate a plurality of peak-shaving and frequency modulation loss judgment results.

[0026] Specifically, the peak-shaving and frequency modulation loss judgment module 14 compares the multiple peak-shaving and frequency modulation loss coefficients calculated by the peak-shaving and frequency modulation loss evaluation module 13 with the preset peak-shaving and frequency modulation loss threshold, and generates a corresponding peak-shaving and frequency modulation loss judgment result according to the comparison result.

[0027] First, the peak-shaving and frequency regulation loss judgment module 14 receives multiple peak-shaving and frequency regulation loss coefficients transmitted by the peak-shaving and frequency regulation loss evaluation module 13. These peak-shaving and frequency regulation loss coefficients correspond one-to-one to each peak-shaving and frequency regulation strategy generated by the peak-shaving and frequency regulation decision module 12, reflecting the estimated loss level of each strategy. Then, the peak-shaving and frequency regulation loss judgment module 14 reads the preset peak-shaving and frequency regulation loss threshold from the storage unit. This threshold represents the upper limit of the acceptable loss of the peak-shaving and frequency regulation strategy, which is set according to the actual operation requirements and experience data of the power grid. Next, the peak-shaving and frequency regulation loss judgment module 14 compares each peak-shaving and frequency regulation loss coefficient with the peak-shaving and frequency regulation loss threshold. For a certain peak-shaving and frequency regulation loss coefficient, if its value is less than the peak-shaving and frequency regulation loss threshold, it indicates that the estimated loss level of the peak-shaving and frequency regulation strategy corresponding to the peak-shaving and frequency regulation loss coefficient is within an acceptable range. At this time, the peak-shaving and frequency regulation loss judgment module 14 generates a judgment result of "acceptable peak-shaving and frequency regulation loss". Conversely, if the value of a peak-shaving and frequency modulation loss coefficient is greater than or equal to the peak-shaving and frequency modulation loss threshold, it means that the estimated loss level of the peak-shaving and frequency modulation strategy corresponding to the peak-shaving and frequency modulation loss coefficient exceeds the acceptable range. In this case, the peak-shaving and frequency modulation loss determination module 14 generates a judgment result of "peak-shaving and frequency modulation loss is unacceptable."

[0028] Through the peak-shaving and frequency regulation loss judgment module 14, a threshold comparison is performed on each peak-shaving and frequency regulation loss coefficient to obtain a peak-shaving and frequency regulation loss judgment result corresponding to each peak-shaving and frequency regulation strategy, providing a decision basis for subsequent peak-shaving and frequency regulation strategy optimization and adjustment.

[0029] The peak-shaving and frequency regulation optimization module 15 is configured to optimize and adjust the multiple peak-shaving and frequency regulation strategies based on the multiple peak-shaving and frequency regulation loss judgment results to generate a peak-shaving and frequency regulation optimization strategy set.

[0030] Specifically, first, the peak-shaving and frequency regulation optimization module 15 receives the peak-shaving and frequency regulation loss judgment result corresponding to each peak-shaving and frequency regulation strategy transmitted by the peak-shaving and frequency regulation loss judgment module 14. For the strategy whose peak-shaving and frequency regulation loss judgment result is "acceptable peak-shaving and frequency regulation loss", it is directly included in the peak-shaving and frequency regulation optimization strategy set; for the strategy whose peak-shaving and frequency regulation loss judgment result is "unacceptable peak-shaving and frequency regulation loss", it enters the optimization adjustment link. In the optimization adjustment link, the peak-shaving and frequency regulation optimization module 15 analyzes the main reasons for the excessively high loss of the peak-shaving and frequency regulation strategy, focusing on the influencing factors such as power quality, power generation resource utilization, equipment wear and economic benefits. Targetedly select the adjustable parameters in the strategy, such as the output adjustment amplitude, rate and time of the second power generation side, and perform local optimization on them to reduce the overall loss of the strategy. The adjusted strategy will be evaluated again by the peak-shaving and frequency regulation loss evaluation module 13 and the peak-shaving and frequency regulation loss judgment module 14. If the peak-shaving and frequency regulation loss judgment result of the strategy changes to "peak-shaving and frequency regulation loss is acceptable", it will be included in the peak-shaving and frequency regulation optimization strategy set; if it is still "peak-shaving and frequency regulation loss is unacceptable", iterative optimization will continue until an acceptable strategy is obtained or the upper limit of the number of iterations is reached.

[0031] Through the optimization and regulation process, the peak-shaving and frequency regulation optimization module 15 continuously improves the existing strategy, minimizing various implementation costs while ensuring safe and stable grid operation, thereby achieving an optimal balance between peak-shaving and frequency regulation effectiveness and economic efficiency. The peak-shaving and frequency regulation optimization module 15 then aggregates all acceptable peak-shaving and frequency regulation strategies found after optimization to form a peak-shaving and frequency regulation optimization strategy set. This provides high-quality, optimal solutions for grid peak-shaving and frequency regulation execution, helping to improve the flexibility and overall efficiency of grid peak-shaving and frequency regulation, and making grid dispatching and operation more intelligent and optimized.

[0032] The peak-shaving and frequency regulation execution module 16 is used to perform peak-shaving and frequency regulation on the power grid based on the multiple peak-gap time zones and according to the second power generation side and the peak-shaving and frequency regulation optimization strategy set.

[0033] Specifically, first, the peak-shaving and frequency regulation execution module 16 receives multiple peak-gap time zones identified by the peak-gap prediction module 11. These time zones are the time periods when the power grid predicts that a power supply gap will occur. At the same time, the peak-shaving and frequency regulation execution module 16 receives the peak-shaving and frequency regulation optimization strategy set generated by the peak-shaving and frequency regulation optimization module 15. The strategy set contains a series of optimized and feasible peak-shaving and frequency regulation strategies. Then, the peak-shaving and frequency regulation execution module 16 monitors in real time whether the current moment is in a peak-gap time zone according to the actual power grid operation status. Once it is detected that the current moment enters the peak-gap time zone, the peak-shaving and frequency regulation execution module 16 selects the corresponding peak-shaving and frequency regulation strategy from the peak-shaving and frequency regulation optimization strategy set, and sends the peak-shaving and frequency regulation task instruction specified in the strategy to the second power generation side. After receiving the peak-shaving and frequency regulation task instruction, the second power generation side adjusts its own power generation output in a timely manner according to the instruction requirements, increases the power supply support to the power grid, and makes up for the power supply shortage in the peak-gap time zone. After the peak-shaving and frequency regulation tasks in the current peak-gap time zone are completed, the peak-shaving and frequency regulation execution module 16 continues to monitor the subsequent peak-gap time zones, and selects the corresponding peak-shaving and frequency regulation strategies for execution according to the specific conditions of each time zone until the peak-shaving and frequency regulation tasks in all peak-gap time zones are completed.

[0034] Through the peak-shaving and frequency regulation execution module 16, the corresponding optimal peak-shaving and frequency regulation strategy is selected according to the real-time operating status of the power grid and the peak-shaving and frequency regulation requirements, and through coordination with the second power generation side, the peak-shaving and frequency regulation tasks of the power grid are efficiently completed to ensure the safe and stable operation of the power grid.

[0035] Furthermore, the embodiment of the present application also includes:

[0036] According to the future time zone window, the load power of the power consumption side of the power grid is predicted, and a predicted load power curve is established; according to the future time zone window, the power generation power of the first power generation side is predicted, and a predicted power generation power curve is established; based on the predicted load power curve, the peak gap characteristics of the predicted power generation power curve are identified to generate the multiple peak gap time zones and the multiple peak power generation gaps.

[0037] In a feasible implementation, first, the peak gap prediction module 11 obtains the time zone window information of the power grid for a period of time in the future as the future time zone window. The future time zone window is a preset, fixed-length future time interval, such as the next 24 hours, the next 7 days, etc. The selection of the time zone window needs to comprehensively consider factors such as the load characteristics of the power grid, the regulation capacity of the peak-shaving and frequency-regulating resources, etc., in order to achieve the best prediction effect. After determining the future time zone window, the peak gap prediction module 11 performs power prediction on the power consumption side and the first power generation side of the power grid respectively. For the power consumption side, historical load data of the power consumption side is collected, such as power consumption and load curves in different time periods, and then the load prediction algorithm is used to predict the load power within the future time zone window. For example, time series analysis method, artificial neural network method, etc., identify the rules and trends of load changes through mining and analysis of historical data, and extrapolate to the future time zone window to obtain a predicted load power curve. For the first power generation side, the peak gap prediction module 11 collects the historical operating data of the generator set, such as the generated power, output regulation records, etc., and comprehensively considers the static parameters such as the type, capacity, efficiency, etc. of the unit, as well as the dynamic influencing factors such as weather conditions and electricity price policies, and uses the prediction algorithm to obtain the predicted power generation curve of the first power generation side in the future time zone window.

[0038] After obtaining the predicted load power curve and predicted generation power curve, the peak gap prediction module 11 performs an overlapping comparative analysis of the two curves. By calculating the power difference between the two curves at each time point, it determines which times present a power gap risk, that is, when the predicted load power exceeds the predicted generation power. Multiple times at which power gap risks occur consecutively are divided into a peak gap time zone, and the power gap values ​​at each time point within this time zone are accumulated to form the peak generation gap for that time zone.

[0039] By analyzing the entire future time zone window, the peak gap prediction module 11 identifies multiple peak gap time zones and corresponding peak power generation gaps, forming a complete power grid power supply gap prediction result, so that the power grid peak and frequency regulation can take targeted control measures in advance, thereby maximizing the safe and stable operation of the power grid.

[0040] Further, such as Figure 2 As shown, the embodiment of the present application also includes:

[0041] Collect the power generation resource parameters of the second power generation side to obtain a peak-shaving and frequency regulation resource set; extract the a-th peak gap time zone and the a-th peak power generation gap according to the multiple peak gap time zones and the multiple peak power generation gaps, where a is a positive integer; perform fusion learning based on P peak-shaving and frequency regulation decision learning models, build a peak-shaving and frequency regulation decision maker, and embed the peak-shaving and frequency regulation decision maker into the peak-shaving and frequency regulation decision module, where P is a positive integer greater than 1; input the peak-shaving and frequency regulation resource set, the a-th peak gap time zone and the a-th peak power generation gap into the peak-shaving and frequency regulation decision maker to obtain the a-th peak-shaving and frequency regulation strategy, and add the a-th peak-shaving and frequency regulation strategy to the multiple peak-shaving and frequency regulation strategies.

[0042] In a preferred embodiment, the peak-shaving and frequency regulation decision module 12 first collects the generation resource parameters of the second generation side, including static and dynamic indicators such as the type, capacity, regulation capability, response speed, and output upper and lower limits of the generator sets on the second generation side. These parameters comprehensively characterize the generation characteristics and peak-shaving and frequency regulation capabilities of the second generation side, forming a complete set of peak-shaving and frequency regulation resources. The peak-shaving and frequency regulation resource set is an important basis for formulating peak-shaving and frequency regulation strategies, determining the extent to which the second generation side can support peak-shaving and frequency regulation of the power grid. Secondly, the peak-shaving and frequency regulation decision module 12 receives multiple peak-shaving time zones and multiple peak-generation gap data transmitted by the peak-shaving gap prediction module 11. For each peak-shaving gap time zone, the peak-shaving and frequency regulation decision module 12 extracts the corresponding time zone and gap data, forming an independent decision task. The multiple peak gaps are numbered using the variable a. The time zone corresponding to the a-th peak gap is the a-th peak gap time zone, and the corresponding generation gap is the a-th peak-generation gap.

[0043] Next, the peak-shaving and frequency regulation decision module 12 uses P peak-shaving and frequency regulation decision learning models to train and learn historical peak-shaving and frequency regulation data. Among them, the peak-shaving and frequency regulation decision learning model adopts common machine learning algorithms such as neural networks and decision trees, and summarizes the inherent laws and experience of peak-shaving and frequency regulation strategy formulation through mining and analysis of historical data. Then, P peak-shaving and frequency regulation decision learning models are used for fusion learning to integrate the advantages of different models and improve the accuracy and generalization ability of strategy decisions. The peak-shaving and frequency regulation decision maker is obtained by fusion learning of the P peak-shaving and frequency regulation decision learning models. The peak-shaving and frequency regulation decision maker can generate corresponding peak-shaving and frequency regulation strategies based on the input peak-shaving and frequency regulation resource status and peak period gap situation. After the peak-shaving and frequency regulation decision maker is built, it will be embedded in the peak-shaving and frequency regulation decision module 12 and become its core functional component. Afterwards, the peak-shaving and frequency regulation decision module 12 inputs the corresponding peak-shaving and frequency regulation resource set, the a-th peak-shaving and frequency regulation time zone, and the a-th peak-shaving and frequency regulation gap into the peak-shaving and frequency regulation decision maker for each peak-shaving gap. After analysis and calculation by the peak-shaving and frequency regulation decision maker, the peak-shaving and frequency regulation strategy for the a-th peak-shaving and frequency regulation gap can be obtained, which is recorded as the a-th peak-shaving and frequency regulation strategy. This strategy will include several specific control measures and implementation plans, such as the output regulation curve of the second power generation side, the reserve capacity mobilization plan, etc. For each peak-shaving and frequency regulation strategy obtained for a peak-shaving and frequency regulation gap, the peak-shaving and frequency regulation decision module 12 adds it to a strategy set to form multiple peak-shaving and frequency regulation strategies.

[0044] Through the peak-shaving and frequency regulation decision module 12, the peak-shaving and frequency regulation resources on the second power generation side are fully utilized, and a feasible peak-shaving and frequency regulation strategy is given for the predicted peak period gap. While ensuring the safety and stability of the power grid, the peak-shaving and frequency regulation benefits of the second power generation side are maximized, so that the supply and demand balance of the power grid can be maintained in a timely and effective manner.

[0045] Furthermore, the embodiment of the present application also includes:

[0046] Based on the power grid, peak-shaving and frequency regulation decision records are collected to obtain peak-shaving and frequency regulation resource set records, peak period gap time zone records, peak period power generation gap records and peak-shaving and frequency regulation strategy records; with the peak-shaving and frequency regulation resource set records, the peak period gap time zone records and the peak period power generation gap records as input information and the peak-shaving and frequency regulation strategy records as output information, supervised training is performed on the P peak-shaving and frequency regulation decision learning models to obtain P frequency regulation decision loss coefficients; it is determined whether the P frequency regulation decision loss coefficients are less than the peak-shaving and frequency regulation decision loss threshold; if the P frequency regulation decision loss coefficients are less than the peak-shaving and frequency regulation decision loss threshold, P peak-shaving and frequency regulation decision units are generated; fusion training is performed based on the P peak-shaving and frequency regulation decision units to generate the peak-shaving and frequency regulation decision maker.

[0047] In a preferred embodiment, the peak-shaving and frequency modulation decision module 12 adopts a fusion learning method based on P peak-shaving and frequency modulation decision learning models when building the peak-shaving and frequency modulation decision maker to improve the performance and generalization ability of the decision maker.

[0048] First, historical records of peak-shaving and frequency regulation decisions are analyzed and collated to obtain data records related to peak-shaving and frequency regulation decisions, including peak-shaving and frequency regulation resource set records, peak-gap time zone records, peak-generation gap records, and peak-shaving and frequency regulation strategy records. The peak-shaving and frequency regulation resource set records reflect the available peak-shaving and frequency regulation resources on the secondary generation side at different times; the peak-gap time zone records and peak-generation gap records reflect various power supply gaps that have occurred in the grid historically; and the peak-shaving and frequency regulation strategy records document the actual control strategies implemented to address these power supply gaps and their effectiveness. This historical data provides valuable samples for training the peak-shaving and frequency regulation decision learning model. After obtaining the historical peak-shaving and frequency regulation decision data, P peak-shaving and frequency regulation decision learning models are trained on this data. Each peak-shaving and frequency regulation decision learning model takes the peak-shaving and frequency regulation resource set records, peak-gap time zone records, and peak-generation gap records as inputs and the peak-shaving and frequency regulation strategy records as the expected outputs. Through supervised learning, the model summarizes the empirical patterns of historical peak-shaving and frequency regulation decisions. Among them, the peak-shaving and frequency-regulating decision learning model can adopt common machine learning algorithms such as neural networks and decision trees, and continuously fit and optimize the mapping relationship between decision input and decision output by adjusting the internal parameters of the model.

[0049] During training, each peak-shaving and frequency modulation decision-making learning model calculates a frequency modulation decision-making loss coefficient to measure the deviation between the decision strategy output by the model and the decision strategy actually executed in historical records. A smaller frequency modulation decision-making loss coefficient indicates that the decision-making rules learned by the model are closer to the actual situation and the decision-making results are better. By evaluating the frequency modulation decision-making loss coefficients of all learning models, the model with the best decision-making performance is selected. Specifically, a peak-shaving and frequency modulation decision-making loss threshold is set, and the frequency modulation decision-making loss coefficients of all peak-shaving and frequency modulation decision-making learning models are evaluated. If a model's frequency modulation decision-making loss coefficient is less than the threshold, it is considered to have learned a relatively accurate and effective decision-making rule and is designated as a qualified peak-shaving and frequency modulation decision-making unit. If a model's frequency modulation decision-making loss coefficient is greater than or equal to the threshold, training continues until its frequency modulation decision-making loss coefficient is less than the threshold. After training, P peak-shaving and frequency modulation decision-making units are obtained. These P peak-shaving and frequency modulation decision-making units are then integrated and trained to generate a peak-shaving and frequency modulation decision-making unit with enhanced overall performance. Fusion training can employ strategies such as weighted averaging and voting to combine the outputs of P peak-shaving and frequency-modulation decision-making units to arrive at the final decision. Because fusion training fully utilizes the collective outputs of these P peak-shaving and frequency-modulation decision-making units, it effectively reduces the limitations and uncertainties of a single model, resulting in a peak-shaving and frequency-modulation decider with greater decision accuracy and stability.

[0050] By performing fusion learning based on P peak-shaving and frequency regulation decision-making learning models, a peak-shaving and frequency regulation decision-maker with superior performance is obtained. It brings together the advantages of multiple machine learning algorithms, can automatically extract peak-shaving and frequency regulation decision-making knowledge from massive historical operation data, and make fast and accurate judgments and strategy generation for new decision-making tasks.

[0051] Furthermore, the embodiment of the present application also includes:

[0052] The peak-shaving and frequency regulation loss evaluation module includes a digital twin unit, a peak-shaving and frequency regulation loss evaluation unit and an initial weight condition for peak-shaving and frequency regulation loss; based on the digital twin unit, the power grid is simulated for peak-shaving and frequency regulation according to the a-th peak-shaving and frequency regulation strategy to obtain the a-th peak-shaving and frequency regulation simulation data set; the a-th peak-shaving and frequency regulation simulation data set is input into the peak-shaving and frequency regulation loss evaluation unit to obtain the a-th peak-shaving and frequency regulation loss evaluation result; based on the a-th peak-shaving and frequency regulation loss evaluation result, a proportion calculation is performed to obtain the a-th peak-shaving and frequency regulation loss improvement weight condition; based on the a-th peak-shaving and frequency regulation loss improvement weight condition, the initial weight condition for peak-shaving and frequency regulation loss is weight-optimized to generate the a-th peak-shaving and frequency regulation loss weight condition; the a-th peak-shaving and frequency regulation loss evaluation result is weightedly calculated according to the a-th peak-shaving and frequency regulation loss weight condition to generate the a-th peak-shaving and frequency regulation loss coefficient, and the a-th peak-shaving and frequency regulation loss coefficient is added to the multiple peak-shaving and frequency regulation loss coefficients.

[0053] In a preferred embodiment, the peak-shaving and frequency regulation loss evaluation module 13 includes a digital twin unit, a peak-shaving and frequency regulation loss evaluation unit, and a peak-shaving and frequency regulation loss initial weight condition. The digital twin unit is a virtual simulation system built based on the physical model of the power grid and historical operating data, which can dynamically simulate the real-time operating status of the power grid. The peak-shaving and frequency regulation loss evaluation unit has a built-in series of quantitative indicators and calculation models for evaluating the losses of the peak-shaving and frequency regulation strategy. The peak-shaving and frequency regulation loss initial weight condition is a set of preset weight parameters used to balance the importance of different loss indicators in the evaluation.

[0054] During the evaluation process, the peak-shaving and frequency regulation loss evaluation module 13 first uses the digital twin unit to simulate and analyze the power grid operation status according to the a-th peak-shaving and frequency regulation strategy. By virtually executing the strategy, the digital twin unit can generate a a-th peak-shaving and frequency regulation simulation data set that is highly consistent with the actual operation, which includes the expected changes in various operating indicators of the power grid after the implementation of the strategy. Then, the peak-shaving and frequency regulation loss evaluation module 13 inputs the a-th peak-shaving and frequency regulation simulation data set into the peak-shaving and frequency regulation loss evaluation unit. The peak-shaving and frequency regulation loss evaluation unit evaluates the execution effect of the strategy from multiple angles based on the built-in quantitative indicators and calculation models of the peak-shaving and frequency regulation strategy losses, and obtains a comprehensive a-th peak-shaving and frequency regulation loss evaluation result, which includes multiple loss indicator values, such as power quality loss, power generation cost loss, equipment loss, etc. Each indicator reflects the degree of impact of the strategy execution on a certain aspect of the power grid performance.

[0055] To rationally determine the weight of each loss indicator in the overall assessment, the peak-shaving and frequency regulation loss evaluation module 13 first calculates the proportion of each loss indicator value to the overall loss based on the first peak-shaving and frequency regulation loss assessment result, thereby obtaining the first peak-shaving and frequency regulation loss weighting condition reflecting the importance of each indicator. The peak-shaving and frequency regulation loss evaluation module 13 then uses the first peak-shaving and frequency regulation loss weighting condition to adaptively adjust the preset initial peak-shaving and frequency regulation loss weighting condition to obtain a more rational first peak-shaving and frequency regulation loss weighting condition. This condition comprehensively considers the emphasis of various indicators under different power grid conditions and can objectively reflect the actual loss level of the peak-shaving and frequency regulation strategy. The peak-shaving and frequency regulation loss evaluation module 13 then applies the first peak-shaving and frequency regulation loss weighting condition to the first peak-shaving and frequency regulation loss assessment result and, through a weighted summation method, calculates a comprehensive first peak-shaving and frequency regulation loss coefficient, which quantitatively describes the overall loss level of the first peak-shaving and frequency regulation strategy. For each obtained peak-shaving and frequency regulation loss coefficient for a strategy, the peak-shaving and frequency regulation loss evaluation module 13 adds it to a loss coefficient set to form multiple peak-shaving and frequency regulation loss coefficients.

[0056] Through the loss assessment process with adaptive weight optimization, the peak-shaving and frequency regulation loss evaluation module 13 can make full use of the digital twin technology of the power grid to efficiently and accurately predict the actual effect of each peak-shaving and frequency regulation strategy in a virtual environment, and dynamically adjust the weight of the loss assessment index according to the real-time status of the power grid, so as to make the assessment results more objective and fair, and provide a reliable decision-making reference for subsequent strategy optimization.

[0057] Furthermore, the embodiment of the present application also includes:

[0058] Data cleaning is performed according to the ath peak-shaving and frequency regulation simulation data set to establish the ath peak-shaving and frequency regulation prediction matrix; the power quality loss assessment model, the power generation resource loss assessment model and the power generation efficiency loss assessment model in the peak-shaving and frequency regulation loss assessment unit are activated; the ath peak-shaving and frequency regulation prediction matrix is ​​input into the power quality loss assessment model to obtain the ath power quality loss assessment coefficient; the ath peak-shaving and frequency regulation prediction matrix is ​​input into the power generation resource loss assessment model to obtain the ath power generation resource loss assessment coefficient; the ath peak-shaving and frequency regulation prediction matrix is ​​input into the power generation efficiency loss assessment model to generate the ath power generation efficiency loss assessment coefficient; the ath power quality loss assessment coefficient, the ath power generation resource loss assessment coefficient and the ath power generation efficiency loss assessment coefficient are output as the ath peak-shaving and frequency regulation loss assessment result.

[0059] In a preferred embodiment, when the peak-shaving and frequency regulation loss evaluation module 13 inputs the a-th peak-shaving and frequency regulation simulation data set into the peak-shaving and frequency regulation loss evaluation unit and obtains the a-th peak-shaving and frequency regulation loss evaluation result. First, the peak-shaving and frequency regulation loss evaluation unit performs data cleaning on the a-th peak-shaving and frequency regulation simulation data set. Data cleaning refers to the process of checking, correcting and filtering the original data, with the aim of improving data quality and eliminating noise, outliers and redundant information in the data. Through data cleaning, the a-th peak-shaving and frequency regulation prediction matrix with standardized format and accurate content is obtained. The matrix contains the expected values ​​of various operating indicators of the power grid after the implementation of the a-th peak-shaving and frequency regulation strategy, which is the basis for subsequent loss evaluation. After obtaining the a-th peak-shaving and frequency regulation strategy, the peak-shaving and frequency regulation loss evaluation unit activates three built-in loss evaluation models, namely the power quality loss evaluation model, the power generation resource loss evaluation model and the power generation efficiency loss evaluation model. These three models quantitatively evaluate the execution effect of the a-th peak-shaving and frequency regulation strategy from the three aspects of power quality, power generation resource utilization and power generation efficiency.

[0060] Subsequently, the peak-shaving and frequency regulation loss assessment unit inputs the a-th peak-shaving and frequency regulation prediction matrix into the power quality loss assessment model to evaluate the impact of the a-th peak-shaving and frequency regulation strategy on grid power quality indicators such as voltage and frequency. By analyzing the changes in relevant indicators in the prediction matrix, a quantitative a-th power quality loss assessment coefficient is derived. The smaller the a-th power quality loss assessment coefficient, the less negative impact on power quality after the strategy is implemented, and the more stable the grid operation. Simultaneously, the peak-shaving and frequency regulation loss assessment unit inputs the a-th peak-shaving and frequency regulation prediction matrix into the power generation resource loss assessment model to evaluate the utilization efficiency and waste of the grid's power generation resources caused by the a-th peak-shaving and frequency regulation strategy. By comparing the changes in indicators such as generator output and spare capacity before and after the implementation of the a-th peak-shaving and frequency regulation strategy, the a-th power generation resource loss assessment coefficient is calculated to measure the decline in power generation resource utilization or waste caused by the strategy's implementation. In addition, the peak-shaving and frequency regulation loss assessment unit inputs the a-th peak-shaving and frequency regulation prediction matrix into the power generation efficiency loss assessment model to evaluate the impact of the a-th peak-shaving and frequency regulation strategy on the power generation equipment and its auxiliary systems. By analyzing the changes in indicators such as the energy consumption level and equipment loss of the power generation equipment after the implementation of the strategy, the a-th power generation efficiency loss assessment coefficient is generated to reflect the problem of decreased power generation efficiency that may be caused by the implementation of the strategy.

[0061] Afterwards, the peak-shaving and frequency regulation loss assessment unit will summarize the ath power quality loss assessment coefficient, the ath power generation resource loss assessment coefficient, and the ath power generation efficiency loss assessment coefficient obtained from the three assessment models, and output the ath peak-shaving and frequency regulation loss assessment result with multiple indicators, which comprehensively reflects the expected losses of the ath peak-shaving and frequency regulation strategy in multiple aspects such as power quality, power generation resource utilization, and power generation efficiency, providing an important reference for subsequent strategy optimization.

[0062] Furthermore, the embodiment of the present application also includes:

[0063] According to the multiple peak-shaving and frequency modulation loss judgment results, the ath peak-shaving and frequency modulation loss judgment result corresponding to the ath peak-shaving and frequency modulation strategy is extracted; if the ath peak-shaving and frequency modulation loss judgment result is that the ath peak-shaving and frequency modulation loss coefficient is less than the peak-shaving and frequency modulation loss threshold, the ath peak-shaving and frequency modulation strategy is output as the ath peak-shaving and frequency modulation optimization strategy; if the ath peak-shaving and frequency modulation loss judgment result is that the ath peak-shaving and frequency modulation loss coefficient is greater than or equal to the peak-shaving and frequency modulation loss threshold, the ath peak-shaving and frequency modulation strategy is optimized and adjusted according to the peak-shaving and frequency modulation loss threshold to generate the ath peak-shaving and frequency modulation optimization strategy; the ath peak-shaving and frequency modulation optimization strategy is added to the peak-shaving and frequency modulation optimization strategy set.

[0064] In a feasible implementation, the peak-shaving and frequency-modulation optimization module 15 adopts a strategy screening method of adaptive threshold optimization to improve the efficiency and accuracy of peak-shaving and frequency-modulation strategy optimization.

[0065] First, the peak-shaving and frequency modulation optimization module 15 extracts the a-th peak-shaving and frequency modulation loss judgment result corresponding to the a-th peak-shaving and frequency modulation strategy based on the multiple peak-shaving and frequency modulation loss judgment results given by the peak-shaving and frequency modulation loss judgment module 14. This judgment result indicates whether the expected loss level of the a-th peak-shaving and frequency modulation strategy exceeds the acceptable range. If the a-th peak-shaving and frequency modulation loss judgment result shows that the a-th peak-shaving and frequency modulation loss coefficient of the a-th peak-shaving and frequency modulation strategy is less than the preset peak-shaving and frequency modulation loss threshold, it means that the loss level of the strategy is low and can be directly used as the preferred strategy. At this time, the peak-shaving and frequency modulation optimization module 15 outputs the a-th peak-shaving and frequency modulation strategy intact as the a-th peak-shaving and frequency modulation optimization strategy, without the need for further optimization and adjustment.

[0066] If the peak-shaving and frequency modulation loss determination result indicates that the peak-shaving and frequency modulation loss coefficient of the a-th peak-shaving and frequency modulation strategy is greater than or equal to the peak-shaving and frequency modulation loss threshold, this indicates that the strategy's loss level is high and there is room for improvement. In this case, the peak-shaving and frequency modulation optimization module 15 optimizes the a-th peak-shaving and frequency modulation strategy based on the peak-shaving and frequency modulation loss threshold. The goal of optimization is to minimize the loss level while ensuring the feasibility of the strategy, ensuring that it meets the peak-shaving and frequency modulation loss threshold. The peak-shaving and frequency modulation optimization module 15 can employ algorithms such as heuristic search and evolutionary optimization to continuously optimize the strategy's loss performance by appropriately adjusting key parameters of the strategy (such as peak-shaving and frequency modulation power and duration) until a peak-shaving and frequency modulation optimization strategy that meets the threshold is generated. Regardless of whether the a-th peak-shaving and frequency modulation strategy is directly output or optimized, the peak-shaving and frequency modulation optimization module 15 ultimately adds the resulting a-th peak-shaving and frequency modulation optimization strategy to the peak-shaving and frequency modulation optimization strategy set, thus compiling all high-quality peak-shaving and frequency modulation strategies.

[0067] Through the adaptive threshold optimization strategy screening process, the peak-shaving and frequency regulation optimization module 15 makes full use of the peak-shaving and frequency regulation loss judgment results to quickly screen and optimize the initial peak-shaving and frequency regulation strategies, obtain high-quality, low-loss optimal strategies, and improve the flexibility and adaptability of power grid peak-shaving and frequency regulation optimization.

[0068] Furthermore, the embodiment of the present application also includes:

[0069] Taking the a-th peak period gap time zone and the a-th peak period power generation gap as the peak-shaving and frequency regulation targets, the a-th peak-shaving and frequency regulation strategy is randomly adjusted according to the peak-shaving and frequency regulation resource set to establish the a-th strategy optimization adjustment space; according to the a-th strategy optimization adjustment space, a first peak-shaving and frequency regulation scheme is randomly extracted; according to the peak-shaving and frequency regulation loss evaluation module, the loss prediction and evaluation of the first peak-shaving and frequency regulation scheme is performed to obtain the peak-shaving and frequency regulation loss coefficient of the first scheme; it is judged whether the peak-shaving and frequency regulation loss coefficient of the first scheme is less than the peak-shaving and frequency regulation loss threshold; if the peak-shaving and frequency regulation loss coefficient of the first scheme is less than the peak-shaving and frequency regulation loss threshold, the first peak-shaving and frequency regulation scheme is added to the a-th peak-shaving and frequency regulation optimization strategy; if the peak-shaving and frequency regulation loss coefficient of the first scheme is greater than or equal to the peak-shaving and frequency regulation loss threshold, the first peak-shaving and frequency regulation scheme is eliminated, and the a-th strategy optimization adjustment space is continued to be optimized and analyzed according to the peak-shaving and frequency regulation loss evaluation module and the peak-shaving and frequency regulation loss threshold until the a-th peak-shaving and frequency regulation optimization strategy is obtained.

[0070] In a preferred embodiment, when the ath peak-shaving and frequency modulation loss judgment result of the ath peak-shaving and frequency modulation strategy shows that the ath peak-shaving and frequency modulation loss coefficient is greater than or equal to the peak-shaving and frequency modulation loss threshold, the peak-shaving and frequency modulation optimization module 15 will further optimize and adjust the ath peak-shaving and frequency modulation strategy to generate the ath peak-shaving and frequency modulation optimization strategy that meets the threshold requirements.

[0071] First, the peak-shaving and frequency regulation optimization module 15 defines the optimization and adjustment objectives, ensuring that the adjusted strategy can fully meet the peak-shaving and frequency regulation needs within the a-peak gap time zone and fill the a-peak power generation gap. With this as the optimization direction, the peak-shaving and frequency regulation optimization module 15 randomly adjusts the key parameters of the a-peak-shaving and frequency regulation strategy based on the grid's currently available peak-shaving and frequency regulation resource set. Based on the strategy's original framework, it forms an optimization and adjustment space for the a-strategy containing multiple possible parameter combinations. After determining the optimization and adjustment space, the peak-shaving and frequency regulation optimization module 15 randomly extracts a new parameter combination from this space to generate a candidate peak-shaving and frequency regulation scheme, denoted as the first peak-shaving and frequency regulation scheme. This scheme retains the basic structure of the a-peak-shaving and frequency regulation strategy, but makes certain adjustments to key parameters to seek to reduce the loss level. To evaluate the actual loss level of the first peak-shaving and frequency regulation scheme, the peak-shaving and frequency regulation optimization module 15 submits it to the peak-shaving and frequency regulation loss evaluation module 13, requesting a loss prediction evaluation. The peak-shaving and frequency regulation loss evaluation module 13 obtains the loss coefficient of the first peak-shaving and frequency regulation scheme, i.e., the peak-shaving and frequency regulation loss coefficient of the first scheme, through digital twin simulation and multi-model evaluation in accordance with the established evaluation process, and feeds it back to the peak-shaving and frequency regulation optimization module 15.

[0072] Subsequently, the peak-shaving and frequency modulation optimization module 15 compares the peak-shaving and frequency modulation loss coefficient of the first scheme with the peak-shaving and frequency modulation loss threshold. If the peak-shaving and frequency modulation loss coefficient of the first scheme is less than the peak-shaving and frequency modulation loss threshold, it means that the loss level of the first peak-shaving and frequency modulation scheme has been effectively controlled and can be used as an ideal optimization result. At this time, the peak-shaving and frequency modulation optimization module 15 will directly add the first peak-shaving and frequency modulation scheme to the a-th peak-shaving and frequency modulation optimization strategy as the final optimization scheme of the a-th peak-shaving and frequency modulation strategy. However, if the peak-shaving and frequency modulation loss coefficient of the first scheme is still greater than or equal to the peak-shaving and frequency modulation loss threshold, it indicates that the loss level of the scheme is not ideal enough and needs further optimization. In this case, the peak-shaving and frequency modulation optimization module 15 temporarily eliminates the first peak-shaving and frequency modulation scheme, and resamples and analyzes the a-th strategy optimization adjustment space based on the gap between the peak-shaving and frequency modulation loss coefficient of the first scheme and the peak-shaving and frequency modulation loss threshold. Through the evaluation capability of the peak-shaving and frequency regulation loss evaluation module 13, the peak-shaving and frequency regulation optimization module 15 continuously extracts new candidate solutions from the a-th strategy optimization adjustment space, evaluates their loss levels, and compares them with the threshold until the optimal solution that meets the loss threshold requirements is found and determined as the a-th peak-shaving and frequency regulation optimization strategy.

[0073] Based on the optimization adjustment process, the peak-shaving and frequency regulation optimization module 15 fully explores the optimization potential of the a-th peak-shaving and frequency regulation strategy, and finds the optimal parameter combination solution through intelligent search and comparison in a given adjustment space, so as to minimize the loss level of the strategy.

[0074] In summary, the power grid peak and frequency regulation adaptive optimization platform provided by the embodiments of the present application has the following technical effects:

[0075] The peak gap prediction module is used to predict peak gaps for the first generation side of the power grid based on future time zone windows, determine multiple peak gap time zones and multiple peak generation gaps, where the first generation side is in grid-connected operation and the second generation side of the power grid is in islanded operation. This module identifies and quantifies the power supply and demand imbalance risk faced by the first generation side of the power grid, providing a key basis for subsequent peak and frequency regulation decisions. The peak and frequency regulation decision module is used to make peak and frequency regulation decisions for the power grid based on the second generation side based on multiple peak gap time zones and multiple peak generation gaps, generate multiple peak and frequency regulation strategies, and propose corresponding peak and frequency regulation countermeasures for different peak gap situations, thereby improving the flexibility and adaptability of the power grid to peak gaps. The peak and frequency regulation loss evaluation module is used to perform loss prediction and evaluation on multiple peak and frequency regulation strategies, obtain multiple peak and frequency regulation loss coefficients, and provide an important reference for subsequent strategy screening and optimization. The peak-shaving and frequency regulation loss judgment module is used to judge whether multiple peak-shaving and frequency regulation loss coefficients are less than the peak-shaving and frequency regulation loss threshold, generate multiple peak-shaving and frequency regulation loss judgment results, judge whether each strategy meets the requirements, and screen out feasible peak-shaving and frequency regulation strategies. The peak-shaving and frequency regulation optimization module is used to optimize and adjust multiple peak-shaving and frequency regulation strategies based on multiple peak-shaving and frequency regulation loss judgment results, and generate a peak-shaving and frequency regulation optimization strategy set. While achieving safe and stable operation of the power grid, it can minimize the overall loss of peak-shaving and frequency regulation. The peak-shaving and frequency regulation execution module is used to perform peak-shaving and frequency regulation on the power grid based on the second power generation side and the peak-shaving and frequency regulation optimization strategy set based on multiple peak gap time zones, to ensure that the power grid can dynamically match the load demand changes when the peak gap occurs, maintain the frequency and voltage stability of the power grid, and ensure the safe operation of the power grid.

[0076] Any step of the platform described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any platform in the embodiments of the present application. No unnecessary restrictions are imposed here.

[0077] Furthermore, the terms "first" or "second" as described above may not only represent an order relationship but may also represent a specific concept and / or refer to the selectability of multiple elements, either individually or in combination. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, if such modifications and variations fall within the scope of this application and its equivalents, this application is intended to include such modifications and variations.

Claims

1. The grid peak and frequency regulation adaptive optimization platform is characterized by: The platform includes: a peak gap prediction module, the peak gap prediction module being configured to perform a peak gap prediction on a first power generation side of a power grid based on a future time zone window, and determine a plurality of peak gap time zones and a plurality of peak power generation gaps, wherein the first power generation side is in a grid-connected operation state and the second power generation side of the power grid is in an islanded operation state; a peak-shaving and frequency regulation decision module, configured to make peak-shaving and frequency regulation decisions on the power grid according to the second power generation side based on the multiple peak-period gap time zones and the multiple peak-period power generation gaps, and generate multiple peak-shaving and frequency regulation strategies; a peak-shaving and frequency-regulation loss evaluation module, configured to perform loss prediction evaluation on the plurality of peak-shaving and frequency-regulation strategies to obtain a plurality of peak-shaving and frequency-regulation loss coefficients; a peak-shaving and frequency-modulation loss judgment module, configured to judge whether the plurality of peak-shaving and frequency-modulation loss coefficients are less than a peak-shaving and frequency-modulation loss threshold, and generate a plurality of peak-shaving and frequency-modulation loss judgment results; A peak-shaving and frequency regulation optimization module, configured to optimize and adjust the multiple peak-shaving and frequency regulation strategies based on the multiple peak-shaving and frequency regulation loss judgment results to generate a peak-shaving and frequency regulation optimization strategy set; A peak-shaving and frequency regulation execution module is used to perform peak-shaving and frequency regulation on the power grid based on the multiple peak-gap time zones, according to the second power generation side and the peak-shaving and frequency regulation optimization strategy set.

2. The power grid peak and frequency regulation adaptive optimization platform according to claim 1, characterized in that: The peak gap prediction module is used to perform peak gap prediction on the first power generation side of the power grid according to the future time zone window, and determine multiple peak gap time zones and multiple peak power generation gaps, including: Performing load power forecasting on the power consumption side of the power grid according to the future time zone window, and establishing a forecast load power curve; performing power generation prediction for the first power generation side according to the future time zone window, and establishing a predicted power generation curve; The predicted power generation power curve is subjected to peak gap feature identification based on the predicted load power curve to generate the multiple peak gap time zones and the multiple peak power generation gaps.

3. The power grid peak and frequency regulation adaptive optimization platform according to claim 1, characterized in that: The peak-shaving and frequency regulation decision module is configured to make peak-shaving and frequency regulation decisions on the power grid according to the second power generation side based on the multiple peak gap time zones and the multiple peak power generation gaps, and generate multiple peak-shaving and frequency regulation strategies, including: Collecting power generation resource parameters of the second power generation side to obtain a peak-shaving and frequency-regulating resource set; Extracting the a-th peak gap time zone and the a-th peak power generation gap according to the multiple peak gap time zones and the multiple peak power generation gaps, where a is a positive integer; Based on P peak-shaving and frequency modulation decision learning models, fusion learning is performed to build a peak-shaving and frequency modulation decision maker, and the peak-shaving and frequency modulation decision maker is embedded into the peak-shaving and frequency modulation decision module, where P is a positive integer greater than 1; The peak-shaving and frequency regulation resource set, the a-th peak period gap time zone and the a-th peak period power generation gap are input into the peak-shaving and frequency regulation decision maker to obtain the a-th peak-shaving and frequency regulation strategy, and the a-th peak-shaving and frequency regulation strategy is added to the multiple peak-shaving and frequency regulation strategies.

4. The grid peak-shaving and frequency-regulating adaptive optimization platform according to claim 3, characterized in that: Based on P peak-shaving and frequency-modulation decision-making learning models, a peak-shaving and frequency-modulation decision-maker is built, including: Collect peak-shaving and frequency regulation decision records based on the power grid to obtain peak-shaving and frequency regulation resource set records, peak gap time zone records, peak power generation gap records, and peak-shaving and frequency regulation strategy records; Using the peak-shaving and frequency regulation resource set record, the peak-period gap time zone record, and the peak-period power generation gap record as input information, and the peak-shaving and frequency regulation strategy record as output information, respectively, supervised training is performed on the P peak-shaving and frequency regulation decision learning models to obtain P frequency regulation decision loss coefficients; Determining whether the P frequency regulation decision loss coefficients are less than a peak regulation and frequency regulation decision loss threshold; If the P frequency regulation decision loss coefficients are less than the peak-shaving and frequency regulation decision loss threshold, generating P peak-shaving and frequency regulation decision units; Fusion training is performed on the P peak-shaving and frequency-modulation decision units to generate the peak-shaving and frequency-modulation decision maker.

5. The grid peak-shaving and frequency-regulating adaptive optimization platform according to claim 1, characterized in that: The peak-shaving and frequency regulation loss evaluation module is used to perform loss prediction evaluation on the multiple peak-shaving and frequency regulation strategies to obtain multiple peak-shaving and frequency regulation loss coefficients, including: The peak-shaving and frequency-regulation loss evaluation module includes a digital twin unit, a peak-shaving and frequency-regulation loss evaluation unit, and an initial weight condition for peak-shaving and frequency-regulation losses; Based on the digital twin unit, simulate peak-shaving and frequency regulation on the power grid according to the a-th peak-shaving and frequency regulation strategy to obtain the a-th peak-shaving and frequency regulation simulation data set; Inputting the a-th peak-shaving and frequency-modulation simulation data set into the peak-shaving and frequency-modulation loss assessment unit to obtain the a-th peak-shaving and frequency-modulation loss assessment result; Calculate the proportion based on the a-th peak-shaving and frequency-regulation loss assessment result to obtain the a-th peak-shaving and frequency-regulation loss increase weight condition; Optimizing the weight configuration of the initial peak-shaving and frequency-modulation loss weight condition based on the a-th peak-shaving and frequency-modulation loss enhancement weight condition to generate the a-th peak-shaving and frequency-modulation loss weight condition; The ath peak-shaving and frequency modulation loss evaluation result is weightedly calculated according to the ath peak-shaving and frequency modulation loss weight condition to generate the ath peak-shaving and frequency modulation loss coefficient, and the ath peak-shaving and frequency modulation loss coefficient is added to the multiple peak-shaving and frequency modulation loss coefficients.

6. The grid peak-shaving and frequency-regulating adaptive optimization platform according to claim 5, characterized in that: Inputting the a-th peak-shaving and frequency-modulation simulation data set into the peak-shaving and frequency-modulation loss assessment unit to obtain the a-th peak-shaving and frequency-modulation loss assessment result, including: Performing data cleaning based on the a-th peak-shaving and frequency-modulation simulation data set to establish the a-th peak-shaving and frequency-modulation prediction matrix; activating a power quality loss assessment model, a power generation resource loss assessment model, and a power generation efficiency loss assessment model in the peak-shaving and frequency-regulating loss assessment unit; Inputting the ath peak-shaving and frequency-regulating prediction matrix into the power quality loss assessment model to obtain the ath power quality loss assessment coefficient; Inputting the ath peak-shaving and frequency-regulating prediction matrix into the power generation resource loss assessment model to obtain the ath power generation resource loss assessment coefficient; Inputting the a-th peak-shaving and frequency-regulating prediction matrix into the power generation efficiency loss assessment model to generate the a-th power generation efficiency loss assessment coefficient; The ath power quality loss assessment coefficient, the ath power generation resource loss assessment coefficient and the ath power generation efficiency loss assessment coefficient are output as the ath peak and frequency regulation loss assessment result.

7. The grid peak-shaving and frequency-regulating adaptive optimization platform according to claim 1, characterized in that: The peak shaving and frequency regulation optimization module is used to optimize and adjust the multiple peak shaving and frequency regulation strategies based on the multiple peak shaving and frequency regulation loss judgment results to generate a peak shaving and frequency regulation optimization strategy set, including: Extracting the a-th peak-shaving and frequency regulation loss judgment result corresponding to the a-th peak-shaving and frequency regulation strategy according to the multiple peak-shaving and frequency regulation loss judgment results; If the a-th peak-shaving and frequency-modulation loss judgment result is that the a-th peak-shaving and frequency-modulation loss coefficient is less than the peak-shaving and frequency-modulation loss threshold, outputting the a-th peak-shaving and frequency-modulation strategy as the a-th peak-shaving and frequency-modulation optimization strategy; If the a-th peak-shaving and frequency-modulation loss judgment result is that the a-th peak-shaving and frequency-modulation loss coefficient is greater than or equal to the peak-shaving and frequency-modulation loss threshold, optimizing and adjusting the a-th peak-shaving and frequency-modulation strategy according to the peak-shaving and frequency-modulation loss threshold to generate the a-th peak-shaving and frequency-modulation optimization strategy; Add the ath peak-shaving and frequency-regulation optimization strategy to the peak-shaving and frequency-regulation optimization strategy set.

8. The grid peak-shaving and frequency-regulating adaptive optimization platform according to claim 7, characterized in that: If the a-th peak shaving and frequency modulation loss judgment result is that the a-th peak shaving and frequency modulation loss coefficient is greater than or equal to the peak shaving and frequency modulation loss threshold, optimizing and adjusting the a-th peak shaving and frequency modulation strategy according to the peak shaving and frequency modulation loss threshold to generate the a-th peak shaving and frequency modulation optimization strategy includes: Taking the a-th peak period gap time zone and the a-th peak period power generation gap as the peak-shaving and frequency regulation targets, randomly adjusting the a-th peak-shaving and frequency regulation strategy according to the peak-shaving and frequency regulation resource set, and establishing the a-th strategy optimal regulation space; According to the a-th strategy, the optimization adjustment space is searched and the first peak and frequency regulation scheme is randomly extracted; Perform loss prediction evaluation on the first peak-shaving and frequency-regulation scheme according to the peak-shaving and frequency-regulation loss evaluation module to obtain a peak-shaving and frequency-regulation loss coefficient of the first scheme; Determining whether the peak-shaving and frequency-regulation loss coefficient of the first scheme is less than the peak-shaving and frequency-regulation loss threshold; If the peak-shaving and frequency-regulation loss coefficient of the first scheme is less than the peak-shaving and frequency-regulation loss threshold, adding the first peak-shaving and frequency-regulation scheme to the ath peak-shaving and frequency-regulation optimization strategy; If the peak-shaving and frequency regulation loss coefficient of the first scheme is greater than or equal to the peak-shaving and frequency regulation loss threshold, the first peak-shaving and frequency regulation scheme is eliminated, and the optimization analysis of the a-th strategy optimization adjustment space is continued according to the peak-shaving and frequency regulation loss evaluation module and the peak-shaving and frequency regulation loss threshold until the a-th peak-shaving and frequency regulation optimization strategy is obtained.

Images