A data-driven virtual power plant optimization scheduling method and system

Abstract

The application provides a kind of data-driven virtual power plant optimization scheduling method and system, the method comprises: obtaining the response adjustment index system of virtual power plant;The index data is obtained by preprocessing the response adjustment index system;Based on the transformation of index data by data-driven method, factor variable is obtained;The factor variable is comprehensively weighted and the virtual power plant comprehensive evaluation score is calculated;Based on the preset reference index and virtual power plant comprehensive evaluation score, the virtual power plant is optimized and scheduled.The data-driven virtual power plant optimization scheduling method proposed in the application, by data-driven method to the index data of virtual power plant is transformed, the objective quantification of virtual power plant scheduling data is realized, the scheduling process of virtual power plant is optimized, and the response success rate of virtual power plant is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual power plant scheduling, and in particular to a data-driven virtual power plant optimization scheduling method and system. Background Art

[0002] Virtual power plants are a key technology for implementing smart distribution networks. They combine distributed clean energy, controllable loads, and energy storage systems within a distribution network into a single, dedicated power plant through a distributed energy management system, effectively reconciling the conflict between smart grids and distributed energy. However, the sheer volume and complexity of metrics in virtual grid scheduling make it difficult to scientifically categorize and comprehensively weight each metric, hindering efficient virtual power plant scheduling.

[0003] Prior art discloses a multidimensional value assessment method and scheduling strategy for demand-side virtual power plants. This method proposes a comprehensive benefit evaluation system for virtual power plant projects based on four dimensions: power savings, investment benefits, economic benefits, and environmental benefits. It also constructs an entropy-fuzzy comprehensive evaluation model to compare and analyze the performance of each indicator to optimize virtual power plant scheduling. Prior art also discloses a comprehensive benefit evaluation method for virtual power plants. This method proposes a comprehensive evaluation index system based on seven indicators, including reliability, economy, and dispatchability, to assess the characteristics and functions of load-based virtual power plants. This system verifies the performance of virtual power plant scheduling in different scenarios. While prior art approaches address the comprehensive evaluation and scheduling of virtual power plants to varying degrees through different approaches, they are subject to significant subjectivity in the classification of indicator data and rely heavily on the evaluator's experience, failing to objectively evaluate the scheduling effectiveness of virtual power plants. Consequently, this method results in a low success rate for virtual power plant responses. Summary of the Invention

[0004] The present invention aims to provide a data-driven virtual power plant optimization scheduling method and system to solve the above-mentioned technical problems. By transforming the index data of the virtual power plant through a data-driven method, the objective quantification of the virtual power plant scheduling data can be achieved, the scheduling process of the virtual power plant can be optimized, and the response success rate of the virtual power plant can be effectively improved.

[0005] In order to solve the above technical problems, the present invention provides a data-driven virtual power plant optimization scheduling method, comprising the following steps:

[0006] Obtain the response and regulation indicator system of the virtual power plant;

[0007] Preprocess the response adjustment index system to obtain index data;

[0008] Transform indicator data based on data-driven methods to obtain factor variables;

[0009] Comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant;

[0010] The virtual power plant is optimized and dispatched based on the preset benchmark indicators and the comprehensive evaluation score of the virtual power plant.

[0011] The above scheme transforms the index data of the virtual power plant through a data-driven method, realizes the objective quantification of the virtual power plant scheduling data, optimizes the scheduling process of the virtual power plant, and effectively improves the response success rate of the virtual power plant.

[0012] Furthermore, the preprocessing of the response adjustment index system to obtain index data specifically includes: performing data verification and standardization processing on the response adjustment index system to obtain index data.

[0013] In the above scheme, by performing data verification and standardization on the response adjustment index system, the common characteristics of the data in the response adjustment index system can be verified and the data dimensions in the response adjustment index system can be eliminated to improve the calculation effect of subsequent data.

[0014] Furthermore, the data-driven method is used to transform the indicator data to obtain factor variables. Specifically, the data-driven method is specifically expressed as:

[0015] αX=α(AF+ε)

[0016] In the formula, α represents the indicator effectiveness matrix; X is the individual indicator vector in the indicator data; A is the loading matrix, and solving the corresponding loading matrix is ​​used to ensure the high interpretability of the factor variable; F is the factor variable, and ε is the interference variable;

[0017] The indicator data is transformed through data-driven methods to obtain factor variables.

[0018] In the above scheme, the data-driven method can aggregate the complex and numerous virtual power plant response adjustment index data into a few independent variables through dimensionality reduction and other methods without losing or minimizing the original index data information.

[0019] Furthermore, the factor variables are comprehensively weighted and the comprehensive evaluation score of the virtual power plant is calculated, specifically:

[0020] The factor variables are comprehensively weighted through the comprehensive weighting model, which is specifically expressed as follows:

[0021]

[0022] Where W n,irepresents the comprehensive weight of the nth virtual power plant with respect to the i-th factor variable; WO n,i represents the objective weight of the nth virtual power plant with respect to the i-th factor variable; WS i represents the subjective weight of the i-th factor variable; m represents the number of all factor variables;

[0023] The comprehensive evaluation score of the virtual power plant is calculated based on the factor variables after comprehensive weighting.

[0024] Furthermore, the optimizing scheduling of the virtual power plant based on the preset benchmark indicators and the comprehensive evaluation score of the virtual power plant includes optimizing scheduling of the virtual power plant scheduling priority and optimizing scheduling of the scheduling capacity.

[0025] The data-driven virtual power plant optimization scheduling method proposed in the above scheme can make full use of the objective and actual virtual power plant operation data, build a rich multi-type indicator evaluation system, and conduct a more scientific and highly interpretable objective evaluation of the virtual power plant scheduling effect, thereby optimizing the virtual power plant scheduling process and improving the response success rate of the virtual power plant.

[0026] The present invention also proposes a data-driven virtual power plant optimization scheduling system for implementing a data-driven virtual power plant optimization scheduling method, which includes a data acquisition module, a preprocessing module, a data conversion module, a comprehensive evaluation score calculation module and an optimization scheduling module; wherein:

[0027] The data acquisition module is used to obtain the response adjustment index system of the virtual power plant;

[0028] The preprocessing module is used to preprocess the response adjustment index system to obtain index data;

[0029] The data transformation module is used to transform the indicator data based on the data-driven method to obtain factor variables;

[0030] The comprehensive evaluation score calculation module is used to comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant;

[0031] The optimization scheduling module is used to optimize the scheduling of the virtual power plant based on preset benchmark indicators and the comprehensive evaluation score of the virtual power plant.

[0032] Furthermore, the preprocessing module is used to preprocess the response adjustment index system to obtain index data. Specifically, the preprocessing module performs data verification and standardization processing on the response adjustment index system to obtain index data.

[0033] Furthermore, the data transformation module is used to transform the indicator data based on a data-driven method to obtain factor variables, specifically:

[0034] The data-driven approach is represented as:

[0035] αX=α(AF+ε)

[0036] In the formula, α represents the indicator effectiveness matrix; X is the individual indicator vector in the indicator data; A is the loading matrix, and solving the corresponding loading matrix is ​​used to ensure the high interpretability of the factor variable; F is the factor variable, and ε is the interference variable;

[0037] The indicator data is transformed through data-driven methods to obtain factor variables.

[0038] Furthermore, the comprehensive evaluation score calculation module is used to comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant, specifically:

[0039] The factor variables are comprehensively weighted through the comprehensive weighting model, which is specifically expressed as follows:

[0040]

[0041] Where W n,i represents the comprehensive weight of the nth virtual power plant with respect to the i-th factor variable; WO n,i represents the objective weight of the nth virtual power plant with respect to the i-th factor variable; WS i represents the subjective weight of the i-th factor variable; m represents the number of all factor variables;

[0042] The comprehensive evaluation score of the virtual power plant is calculated based on the factor variables after comprehensive weighting.

[0043] Furthermore, the optimization scheduling module is used to optimize the scheduling of the virtual power plant based on preset benchmark indicators and the comprehensive evaluation score of the virtual power plant, including: optimizing the scheduling priority of the virtual power plant and optimizing the scheduling capacity. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A flowchart of a data-driven virtual power plant optimization scheduling method provided by one embodiment of the present invention;

[0045] Figure 2 A schematic diagram of a response adjustment index system provided by one embodiment of the present invention;

[0046] Figure 3 A module connection diagram of a data-driven virtual power plant optimization scheduling system provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0048] See Figure 1 This embodiment provides a data-driven virtual power plant optimization scheduling method, including the following steps:

[0049] S1: Obtain the response regulation indicator system of the virtual power plant;

[0050] S2: pre-process the response adjustment index system and obtain index data;

[0051] S3: Transform the indicator data based on the data-driven method to obtain factor variables;

[0052] S4: Comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant;

[0053] S5: Optimize the scheduling of virtual power plants based on preset benchmark indicators and comprehensive evaluation scores of virtual power plants.

[0054] Further, see Figure 2 The response adjustment index system includes time index data, power index data, economic index data and other index data.

[0055] Time-related indicator data refers to the time-related indicator data during the scheduling execution of the virtual power plant, including but not limited to: regulation start time, regulation ramp time, regulation recovery time and minimum continuous downtime, etc.

[0056] The adjustment start time is the time interval from when the virtual power plant receives the dispatch instruction to when it actually starts to execute the dispatch. It is used to characterize the response speed of the virtual power plant to the dispatch instruction, which can be specifically expressed as: T 启动 =t s -t c , where t s is the actual adjustment start time of the virtual power plant, t c The time when the virtual power plant receives adjustment instructions.

[0057] The regulation ramp time is the time interval from the start of the virtual power plant to the entry into the effective regulation range. It represents the speed at which the virtual power plant enters the effective regulation state and reflects its regulation flexibility and capability. It can be specifically expressed as: T 爬坡 =t v -t s , where tv The time when the virtual power plant enters the effective regulation range.

[0058] The regulation recovery time is the time interval from when the virtual power plant leaves the effective regulation range due to disturbance to when it re-enters the effective regulation range. It represents the speed at which the virtual power plant eliminates the disturbance and returns to normal, reflecting its anti-interference ability. It can be specifically expressed as T 恢复 =t r -t d , where t r is the time it takes for the virtual power plant to re-enter the effective regulation range after the disturbance, t d The time when the virtual power plant is disturbed and out of the effective regulation range.

[0059] The minimum continuous downtime is the minimum downtime interval between two independent regulation instructions of the virtual power plant, which represents the speed at which the virtual power plant can resume its regulation execution capability after executing one regulation. express.

[0060] Power index data refers to the power-related index data of the virtual power plant during the scheduling execution process, including but not limited to: maximum regulation capacity, minimum regulation capacity, maximum up-ramp rate and maximum down-ramp rate, etc.

[0061] The maximum regulation capacity is the maximum power difference of the virtual power plant to perform independent stable regulation, which represents the maximum capacity of the virtual power plant to maintain stable power regulation. It can be divided into the maximum upward regulation capacity and the maximum downward regulation capacity. and express.

[0062] The minimum regulation capability is the lowest power difference for the virtual power plant to perform independent stable regulation, which represents the starting ability of the virtual power plant to maintain stable power regulation. It is similar to the concept of the minimum technical output of thermal power units and can be divided into the minimum upward regulation capability and the minimum downward regulation capability. and express.

[0063] The maximum ramp rate is the maximum power increase per unit time when the virtual power plant performs independent regulation, which represents the ability of the virtual power plant to adjust the power change speed upward. It can be specifically expressed as Among them, P t+1 and P t The real-time power of the virtual power plant in period t+1 and period t respectively.

[0064] The maximum ramp-down rate is the maximum power drop per unit time when the virtual power plant performs independent regulation, which represents the ability of the virtual power plant to adjust the power change speed downward. It can be specifically expressed as

[0065] Economic indicator data refers to economic-related indicator data during the scheduling and execution of virtual power plants, including but not limited to: cumulative revenue, single maximum revenue, historical highest quotation and average quotation, etc.

[0066] Cumulative income refers to the cumulative income of the virtual power plant's regulation business within a certain period of time, which represents the profitability of the virtual power plant's regulation business during that period of time. express.

[0067] The single maximum profit refers to the highest historical profit of the virtual power plant in performing independent regulation business, which represents the profitability of the virtual power plant in participating in single regulation. express.

[0068] The highest historical bid refers to the highest historical bid for independent regulation transactions of virtual power plants within a certain period of time, which represents the upper limit of the bid for virtual power plants to participate in market transactions. express.

[0069] The average bid refers to the historical average bid of the virtual power plant for independent adjustment transactions within a certain period of time, which represents the average bid level of the virtual power plant in the market transaction. express.

[0070] Other indicator data refers to other virtual power plant dispatch execution indicator data that cannot be classified into the previous three indicators, including but not limited to: regulation participation rate, regulation qualification rate, regulation accuracy rate and renewable energy consumption rate.

[0071] The regulation participation rate refers to the ratio of the number of regulation periods in which the virtual power plant participates in demand regulation to the number of historical demand regulation periods within a certain time period. It represents the initiative of the virtual power plant to participate in demand regulation during the time period. It can be specifically expressed as:

[0072]

[0073] where N 参与 Indicates the total number of time periods that the virtual power plant participates in regulation within a certain time period, N 需求 Indicates the total number of demand periods within the time period.

[0074] The regulation qualification rate refers to the ratio of the number of periods in which the virtual power plant meets the regulation standard to the number of historical response periods within a certain period of time. It represents the ability of the virtual power plant to achieve effective regulation and can be specifically expressed as:

[0075]

[0076] where N 合格 It indicates the total number of time periods during which the virtual power plant is regulated to meet the standards within a certain period of time.

[0077] Regulation accuracy refers to the ratio of the number of accurate regulation periods of the virtual power plant to the number of historical response periods within a certain period of time. It represents the ability of the virtual power plant to achieve accurate regulation and can be specifically expressed as:

[0078]

[0079] where N 准确 It indicates the total number of time periods during which the virtual power plant can achieve accurate regulation within a certain time period. The standard for accurate regulation can be determined according to scheduling requirements.

[0080] The renewable energy absorption rate refers to the ratio of renewable energy generation to electricity consumption of a virtual power plant within a certain period of time, which represents the ability of the virtual power plant to absorb renewable energy. It can be specifically expressed as:

[0081]

[0082] where Q 新能源 It represents the total power generation of renewable energy in a certain period of time by the virtual power plant, Q 总用电 Indicates the total electricity consumption of the virtual power plant during the time period.

[0083] Furthermore, the pre-processing of the response adjustment index system to obtain index data specifically includes: performing data verification and standardization processing on the response adjustment index system to obtain index data.

[0084] In this embodiment, the purpose of performing data verification on the response adjustment index system is to perform common feature verification on the data in the response adjustment index system. Specifically, this can be achieved by using a correlation matrix, a KMO test method, or a Bartlett sphericity test method. When the data fails the test, the response adjustment index system is reacquired and adjusted until the data common feature requirements are met.

[0085] In this embodiment, the purpose of standardizing the response adjustment index system is to eliminate the data dimension in the response adjustment index system to improve the calculation effect of subsequent data. Specifically, the following standardized nursing formula can be used for this purpose:

[0086] The standardization formula can be expressed as:

[0087]

[0088] Where: z is the standardized value; x is the observed value of the individual indicator data; μ is the mean of the overall data, and σ is the standard deviation of the overall data.

[0089] Furthermore, the data-driven method is used to transform the indicator data to obtain factor variables. Specifically, the data-driven method is specifically expressed as:

[0090] αX=α(AF+ε)

[0091] In the formula, α represents the indicator effectiveness matrix; X is the individual indicator vector in the indicator data; A is the loading matrix, and solving the corresponding loading matrix is ​​used to ensure the high interpretability of the factor variable; F is the factor variable, and ε is the interference variable;

[0092] The indicator data is transformed through data-driven methods to obtain factor variables.

[0093] It should be noted that the above-mentioned data-driven method is essentially a method of aggregating complex and numerous virtual power plant response adjustment index data into a few independent variables through dimensionality reduction and other means with minimal or no loss of original index data information.

[0094] In this embodiment, the indicator validation matrix is ​​a diagonal matrix. When all indicators are valid, this diagonal matrix becomes the identity matrix. Setting the mth row of the indicator validation matrix to 0 indicates ignoring the mth indicator data in the individual indicator vector in the indicator data. Factor variables F are also called highly interpretable factor variables. Their high interpretability is ensured by solving an appropriate loading matrix A. If the solved factor variable F does not meet the high interpretability requirement, the loading matrix is ​​rotated and the data-driven method is re-solved. Rotation methods such as varimax and quarticmax can be used.

[0095] It should be noted that the solution of the load matrix A is based on the index data. Specifically, it can be solved by principal component analysis, principal factor method, maximum likelihood estimation method, etc. and expressed as A, which is used to substitute into the improved data-driven method for further solution.

[0096] Furthermore, the factor variables are comprehensively weighted and the comprehensive evaluation score of the virtual power plant is calculated, specifically:

[0097] The factor variables are comprehensively weighted through the comprehensive weighting model, which is specifically expressed as follows:

[0098]

[0099] Where W n,i represents the comprehensive weight of the nth virtual power plant with respect to the i-th factor variable; WO n,i represents the objective weight of the nth virtual power plant with respect to the i-th factor variable; WS i represents the subjective weight of the i-th factor variable; m represents the number of all factor variables;

[0100] The comprehensive evaluation score of the virtual power plant is calculated based on the factor variables after comprehensive weighting.

[0101] It should be noted that the comprehensive weight of the nth virtual power plant can be represented by the vector Wn =[W n,1 ,W n,2 ,...,W n,m ]express.

[0102] In this embodiment, the objective weight determination method and subjective weight determination method of the factor variables can be selected according to actual needs. Specifically, this embodiment uses the hierarchical analysis method to determine the subjective weight and uses the calculation of the variance explanation rate of the corresponding factor variables to determine the objective weight.

[0103] In this embodiment, the comprehensive evaluation score of the virtual power plant can be obtained by weighted averaging the factor variables. The specific calculation method can be expressed as follows:

[0104] M=WF

[0105] Where, M=[M1,M2,...,M N ] is the comprehensive evaluation score of N virtual power plants, W=[W1,...,W n ,...,W N ] T is the comprehensive weight matrix corresponding to N virtual power plants, F=[F1,F2,...,F m ] T is a vector of m highly explanatory factor variables.

[0106] Furthermore, the optimizing scheduling of the virtual power plant based on the preset benchmark indicators and the comprehensive evaluation score of the virtual power plant includes optimizing scheduling of the virtual power plant scheduling priority and optimizing scheduling of the scheduling capacity.

[0107] It should be noted that the preset benchmark indicators are obtained by the virtual power plant dispatching center through long-term historical data and dispatching expectations. Specifically, they can be obtained by generating the expected random indicator data set through the Monte Carlo simulation method and calculating various stable benchmark indicators.

[0108] In this embodiment, the essence of optimizing the scheduling of virtual power plants is to prioritize the currently available virtual power plant resources and optimize their scheduling capacity based on the comprehensive evaluation score of the virtual power plant, thereby improving the response success rate of the virtual power plant.

[0109] In this embodiment, the scheduling priority of the virtual power plant can be optimized according to the calculated comprehensive evaluation score of the virtual power plant. After each calculation of the comprehensive evaluation score of the virtual power plant is completed, the priority sorting of the currently available virtual power plant regulation resources is updated, that is, the virtual power plant regulation resources are sorted from high to low according to the comprehensive evaluation score. In subsequent scheduling, the regulation resources with high comprehensive evaluation scores are given priority according to demand.

[0110] The optimization of the virtual power plant dispatch capacity can be performed through the following formula, which is specifically expressed as follows:

[0111]

[0112] Where: C dec Declare capacity for virtual power plants, M VPP is the comprehensive evaluation score of the virtual power plant, M s is the virtual power plant score calculated based on the benchmark indicators, C out It is the actual available capacity of the virtual power plant after optimization.

[0113] The data-driven virtual power plant optimization scheduling method proposed in this embodiment can make full use of objective and actual virtual power plant operation data, construct a rich multi-type indicator evaluation system, and conduct a more scientific and highly interpretable objective evaluation of the virtual power plant scheduling effect, thereby optimizing the virtual power plant scheduling process and improving the response success rate of the virtual power plant.

[0114] See Figure 3 This embodiment proposes a data-driven virtual power plant optimization scheduling system for implementing a data-driven virtual power plant optimization scheduling method, which includes a data acquisition module, a preprocessing module, a data transformation module, a comprehensive evaluation score calculation module, and an optimization scheduling module; wherein:

[0115] The data acquisition module is used to obtain the response adjustment index system of the virtual power plant;

[0116] The preprocessing module is used to preprocess the response adjustment index system to obtain index data;

[0117] The data transformation module is used to transform the indicator data based on the data-driven method to obtain factor variables;

[0118] The comprehensive evaluation score calculation module is used to comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant;

[0119] The optimization scheduling module is used to optimize the scheduling of the virtual power plant based on preset benchmark indicators and the comprehensive evaluation score of the virtual power plant.

[0120] Furthermore, the preprocessing module is used to preprocess the response adjustment index system to obtain index data. Specifically, the preprocessing module performs data verification and standardization processing on the response adjustment index system to obtain index data.

[0121] Furthermore, the data transformation module is used to transform the indicator data based on a data-driven method to obtain factor variables, specifically:

[0122] The data-driven approach is represented as:

[0123] αX=α(AF+ε)

[0124] In the formula, α represents the indicator effectiveness matrix; X is the individual indicator vector in the indicator data; A is the loading matrix, and solving the corresponding loading matrix is ​​used to ensure the high interpretability of the factor variable; F is the factor variable, and ε is the interference variable;

[0125] The indicator data is transformed through data-driven methods to obtain factor variables.

[0126] Furthermore, the comprehensive evaluation score calculation module is used to comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant, specifically:

[0127] The factor variables are comprehensively weighted through the comprehensive weighting model, which is specifically expressed as follows:

[0128]

[0129] Where W n,i represents the comprehensive weight of the nth virtual power plant with respect to the i-th factor variable; WO n,i represents the objective weight of the nth virtual power plant with respect to the i-th factor variable; WS i represents the subjective weight of the i-th factor variable; m represents the number of all factor variables;

[0130] The comprehensive evaluation score of the virtual power plant is calculated based on the factor variables after comprehensive weighting.

[0131] Furthermore, the optimization scheduling module is used to optimize the scheduling of the virtual power plant based on preset benchmark indicators and the comprehensive evaluation score of the virtual power plant, including: optimizing the scheduling priority of the virtual power plant and optimizing the scheduling capacity.

[0132] In this embodiment, since virtual power plants generate a large amount of operating data during their actual operation (such as participating in market transactions), the actual operating data generated is of great value to the formulation of policies and engineering applications related to virtual power plants. This system can implement a data-driven virtual power plant optimization scheduling method, fully exploit the value of operating data, objectively and effectively evaluate the scheduling of virtual power plants, and thus optimize the scheduling process of virtual power plants, which is conducive to guiding virtual power plants to strengthen their control level construction, improve the response success rate of virtual power plants and their level of participation in market operations.

[0133] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A data-driven virtual power plant optimization scheduling method, characterized in that: The following steps are involved: Obtain the response and regulation indicator system of the virtual power plant; Preprocess the response adjustment index system to obtain index data; Based on the data-driven method, the indicator data is transformed to obtain factor variables. Specifically, the data-driven method is specifically expressed as: αX = α(AF + ε); where α represents the indicator effectiveness matrix; X is the individual indicator vector in the indicator data; A is the loading matrix, and solving the corresponding loading matrix is ​​used to ensure the high interpretability of the factor variable; F is the factor variable, and ε is the interference variable; Comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant; specifically, comprehensively weight the factor variables through the comprehensive weighting model, which is specifically expressed as: Where W n,i represents the comprehensive weight of the nth virtual power plant with respect to the i-th factor variable; WO n,i represents the objective weight of the nth virtual power plant with respect to the i-th factor variable; WS i represents the subjective weight of the i-th factor variable; m represents the number of all factor variables; The virtual power plant is optimized and dispatched based on the preset benchmark indicators and the comprehensive evaluation score of the virtual power plant.

2. The data-driven virtual power plant optimization scheduling method according to claim 1 is characterized in that: The preprocessing of the response adjustment index system to obtain index data specifically includes: performing data inspection and standardization processing on the response adjustment index system to obtain index data.

3. The data-driven virtual power plant optimization scheduling method according to claim 1 is characterized in that: The optimizing scheduling of the virtual power plant based on the preset benchmark indicators and the comprehensive evaluation score of the virtual power plant includes optimizing scheduling of the virtual power plant scheduling priority and optimizing scheduling of the scheduling capacity.

4. A data-driven virtual power plant optimization scheduling system, characterized in that: It includes data acquisition module, preprocessing module, data transformation module, comprehensive evaluation score calculation module and optimization scheduling module; among which: The data acquisition module is used to obtain the response adjustment index system of the virtual power plant; The preprocessing module is used to preprocess the response adjustment index system to obtain index data; The data transformation module is used to transform the indicator data based on the data-driven method to obtain factor variables. Specifically, the data-driven method is specifically expressed as: αX = α(AF + ε); where α represents the indicator effectiveness matrix; X is the individual indicator vector in the indicator data; A is the loading matrix, and solving the corresponding loading matrix is ​​used to ensure high interpretability of the factor variables; F is the factor variable, and ε is the interference variable; The comprehensive evaluation score calculation module is used to comprehensively weight the factor variables and calculate the comprehensive evaluation score of the virtual power plant; specifically, the factor variables are comprehensively weighted through a comprehensive weighting model, which is specifically expressed as: Where W n,i represents the comprehensive weight of the nth virtual power plant with respect to the i-th factor variable; WO n,i represents the objective weight of the nth virtual power plant with respect to the i-th factor variable; WS i represents the subjective weight of the i-th factor variable; m represents the number of all factor variables; The optimization scheduling module is used to optimize the scheduling of the virtual power plant based on preset benchmark indicators and the comprehensive evaluation score of the virtual power plant.

5. The data-driven virtual power plant optimization scheduling system according to claim 4 is characterized in that: The preprocessing module is used to preprocess the response adjustment index system to obtain index data. Specifically, the preprocessing module performs data inspection and standardization processing on the response adjustment index system to obtain index data.

6. The data-driven virtual power plant optimization scheduling system according to claim 4 is characterized in that: The optimization scheduling module is used to optimize the scheduling of the virtual power plant based on preset benchmark indicators and the comprehensive evaluation score of the virtual power plant, including: optimizing the scheduling priority of the virtual power plant and optimizing the scheduling capacity.

Images