Method and system for identifying market power risk at two sides of sale for electricity market

Abstract

The invention discloses a method for identifying market power risks on both sides of a power market. The method comprises the following steps: acquiring transaction data information of a power generation side and a power selling side of a target power market; constructing a bilateral matching scheme based on a Borda counting method through a preference sorting mechanism, and performing screening to obtain an abnormal matching pair initial set of power generators and electricity sellers; calculating corresponding joint probability distribution and Shannon entropy; calculating joint entropy and mutual information; and screening high-risk transaction pairs so as to complete identification of market power risks on the two sides of the electricity market. The invention also discloses a system for realizing the market power risk identification method for the two selling sides of the electricity market. According to the method, identification of market power risks on the two sides of sale of the electricity market is realized, the reliability is higher, and the accuracy is better.

Description

Technical Field

[0001] The present invention belongs to the field of electrical automation, and in particular relates to a method and system for identifying market power risks on both sides of the supply chain in an electricity market. Background Art

[0002] With the development of economy and technology and the improvement of people's living standards, electricity has become an indispensable secondary energy source in people's production and life, bringing endless convenience to people's production and life. Therefore, ensuring a stable and reliable supply of electricity has become one of the most important tasks of the power system.

[0003] In recent years, with the rapid development of the electricity market, trading entities have become increasingly diversified. At the same time, abuses of market power have also emerged, such as collusion between power generators and electricity retailers to manipulate prices and restrict supply. These abuses not only hinder the normal operation of the electricity market but also pose a serious threat to the safe and reliable operation of the power system and the stable and reliable supply of electricity.

[0004] At present, some researchers have proposed market power identification schemes involving the electricity market, such as the market power identification scheme with application number 202411738953.8. This scheme first constructs a market power indicator system for the electricity market, covering market concentration, bidding strategies, and collusion behavior characteristics; secondly, by integrating the elastic network regression model of the L1 norm (Lasso regression) and the L2 norm (ridge regression), a penalty function is constructed to screen key indicators, solve the multicollinearity problem and achieve variable sparsity, screen out indicators that have a significant impact on market power, and realize the rapid identification of high-market power power generation entities. However, this scheme only considers the abnormal market power behavior of power generators in the electricity market, and does not consider the market power risks on both sides of the issuance. Summary of the Invention

[0005] One of the objectives of the present invention is to provide a method for identifying market power risks on both sides of the electricity market with high reliability and good accuracy.

[0006] A second object of the present invention is to provide a system for implementing the method for identifying market power risks on both sides of the electricity market.

[0007] The method for identifying market power risks on both sides of the electricity market provided by the present invention comprises the following steps:

[0008] S1. Obtain transaction data information on the power generation and sales sides of the target power market;

[0009] S2. Based on the data information obtained in step S1, a bilateral matching scheme is constructed using a preference ranking mechanism and the Borda counting method to screen out an initial set of abnormal matching pairs between power generators and electricity retailers.

[0010] S3. Calculate the corresponding joint probability distribution and Shannon entropy value based on the initial set of abnormal matching pairs obtained in step S2;

[0011] S4. Calculate the joint entropy and mutual information based on the Shannon entropy obtained in step S3;

[0012] S5. Based on the calculation results obtained in step S4, high-risk transaction pairs are screened to complete the identification of market power risks on both sides of the electricity market.

[0013] Step S2, based on the data information obtained in step S1, constructs a bilateral matching scheme based on the Borda counting method through a preference sorting mechanism, and screens out an initial set of abnormal matching pairs of power generators and power retailers. Specifically, the following steps are included:

[0014] Preference types are set based on the needs of power generators and power retailers. These preference types include bid preference, transaction volume preference, and cooperation stability preference. Bid preference is defined as: ranking the bids of power generators and power retailers, with the highest bidder given priority. Transaction volume preference is defined as: ranking the bids of power generators and power retailers, with the bidder with the most recent bid given priority. Cooperation stability preference is defined as: ranking the bids of power generators and power retailers based on the number of times they have participated in transactions, with the bidder with the most transactions given priority.

[0015] Based on the Borda counting method, the acquired data information is converted into Borda scores:

[0016] For any preferred type of power generator A i To e-commerce seller B j , use the following formula to calculate A i and B j When matching A i Satisfaction

[0017]

[0018] Where n is the total number of electricity sellers; j k is the preference order of electricity sellers for electricity generators;

[0019] Setting: When the generator’s preference order for s retailers is j k When A i and B j When matching A i Satisfaction for

[0020] Then we can get the Borda score matrix under the preference ranking of power generators to electricity retailers:

[0021] For any preferred type of seller B j For power generator A i , use the following formula to calculate A i and B j Matching B j Satisfaction

[0022]

[0023] Where m is the total number of power generators; j l is the preference order of generators for retailers;

[0024] Assume that the order of preference of the seller to the s distributors is j. l When A i and B j Matching B j Satisfaction for

[0025] Then we can get the Borda score matrix under the preference ranking of electricity sellers to power generators:

[0026] Calculate the Borda score matrix S′ under the quotation preference respectively price-A and S′ price-B , Borda score matrix S′ under volume preference volume-A and S′ volume-B And the Borda score matrix S′ under cooperative stability preference stable-A and S′ stable-B ;

[0027] For each Borda score matrix, calculate the corresponding extreme quantile Q 90 , set when the satisfaction is greater than the extreme quantile Q 90 When is marked as abnormal, the corresponding Boolean matrix is ​​obtained:

[0028]

[0029]

[0030] In the formula is the Boolean matrix of price preference on the power generation side; S′ price-A The element in row i and column j of ; Based on S′ price-A Calculated Q 90 extreme quantiles; is the Boolean matrix of price preference on the electricity sales side; S′ price-B The element in row i and column j of ; Based on S′ price-B Calculated Q 90 extreme quantiles; is the Boolean matrix of transaction volume preference on the generation side; S′ volume-A The element in row i and column j of ; Based on S′ volume-A Calculated Q 90 extreme quantiles; is the Boolean matrix of transaction volume preference on the electricity sales side; S′ volume-B The element in row i and column j of ; Based on S′ volume-B Calculated Q 90 extreme quantiles; is the Boolean matrix of cooperation stability preference on the generation side; S′ stable-A The element in row i and column j of ; Based on S′ stable-A Calculated Q 90 extreme quantiles; is the Boolean matrix of cooperation stability preference for the electricity sales side; S′ stable-B The element in row i and column j of ; Based on S′ stable-B Calculated Q 90 extreme quantiles;

[0031] Will and Merge into the power generation side comprehensive label matrix S A , S A The element in row i and column j The value of

[0032]

[0033] Will and Merge into the comprehensive label matrix S on the power sales side B , S B The element in row i and column j The value of

[0034]

[0035] Comprehensive power generation side comprehensive label matrix S A And the comprehensive label matrix S of the power sales side B , we get the matching matrix S of the two sides of the electricity sales, and the element s in the i-th row and j-th column of S is ij The value of

[0036] According to the element values ​​in the matching matrix S on both sides of the power supply, the matching pairs of power generators and power retailers corresponding to the element value of 1 are screened and used as the initial set of abnormal matching pairs of power generators and power retailers F = {(A i ,B j )|s ij =1}.

[0037] Step S3, based on the initial set of abnormal matching pairs obtained in step S2, calculates the corresponding joint probability distribution and Shannon entropy value, which specifically includes the following steps:

[0038] According to the initial set F of abnormal matching pairs obtained in step S2, for any abnormal matching pair (A i ,B j ), get the generator A i and retailer B j Set the historical transaction price and transaction volume data within the historical time period; set the historical transaction price X and transaction volume Y, and record all possible discrete values ​​of X as x k , all possible discrete values ​​of Y are y l ;

[0039] The following formula is used to calculate the transaction price x k and power l Perform preprocessing:

[0040]

[0041]

[0042] In the formula is the standardized value of the transaction price; μ x is the mean of the transaction price; σ x is the standard deviation of the transaction price; is the standardized value of transaction volume; μ y is the mean value of transaction volume; σ y is the standard deviation of transaction volume;

[0043] The joint probability distribution f(x,y) of X and Y is calculated using the following formula:

[0044]

[0045] where M is the number of power generators; N is the number of power sellers; h x is the bandwidth parameter of the Gaussian kernel function of the transaction electricity price; h y is the bandwidth parameter of the Gaussian kernel function of the transaction electricity quantity; K is the Gaussian kernel function;

[0046] The transaction electricity price and the transaction electricity quantity of the power generator A i and the power seller B j are calculated by the following formula:

[0047]

[0048] The Shannon entropy H(X) of the transaction electricity price X is calculated by the following formula:

[0049]

[0050] where P(x k ) is the probability distribution of the transaction electricity price;

[0051] The Shannon entropy H(Y) of the transaction electricity quantity Y is calculated by the following formula:

[0052]

[0053] where Q(y l ) is the probability distribution of the transaction electricity quantity.

[0054] According to the Shannon entropy values obtained in step S3, the joint entropy and the mutual information are calculated, and the calculation specifically includes the following steps:

[0055] According to the obtained joint entropy H(X,Y), the mutual information I(X;Y) between the transaction electricity price and the transaction electricity quantity between the power generator A i and the power seller B j is calculated as I(X;Y) = H(X) + H(Y) - H(X,Y).

[0056] According to the calculation result obtained in step S4, the high-risk transaction pair is screened, and the screening specifically includes the following steps:

[0057] The mutual information threshold is set;

[0058] If the mutual information I(X;Y) between the power generation side A i and the power selling side B j obtained in step S4 is greater than the mutual information threshold, it is determined that the power generation side A i and the power selling side B j are a high-risk transaction pair.

[0059] The application further provides a system for realizing the method for identifying the market power risk of the power market, comprising a data acquisition module, an initial matching module, an entropy value calculation module, a mutual information calculation module and a risk identification module; the data acquisition module, the initial matching module, the entropy value calculation module, the mutual information calculation module and the risk identification module are connected in sequence; the data acquisition module is used for acquiring the transaction data information of the power generation side and the power sale side of a target power market, and uploading the data information to the initial matching module; the initial matching module is used for constructing a double-sided matching scheme based on the Borda counting method through a preference ordering mechanism according to the received data information, screening an initial set of abnormal matching pairs of power suppliers and power sellers, and uploading the data information to the entropy value calculation module; the entropy value calculation module is used for calculating the corresponding joint probability distribution and Shannon entropy value according to the received data information and the obtained initial set of abnormal matching pairs, and uploading the data information to the mutual information calculation module; the mutual information calculation module is used for calculating the joint entropy and mutual information according to the received data information and the obtained Shannon entropy value, and uploading the data information to the risk identification module; and the risk identification module is used for screening high-risk transaction pairs according to the received data information and the obtained calculation result, so as to complete the identification of the market power risk of the power market.

[0060] The system is realized based on the SpringCloud micro-service architecture.

[0061] The method and system for identifying the market power risk of the power market provided by the application realize the identification of the market power risk of the power market through the acquisition, initial matching and statistical parameter calculation and screening of the data information of the power generation side and the power sale side, and have higher reliability and better accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 The figure is a method flowchart of the method of the application.

[0063] Figure 2 The figure is a function module schematic diagram of the system of the application. DETAILED DESCRIPTION

[0064] As Figure 1 The figure is a method flowchart of the method of the application: the method for identifying the market power risk of the power market disclosed by the application comprises the following steps:

[0065] S1. acquiring the transaction data information of the power generation side and the power sale side of a target power market;

[0066] S2. Based on the data information obtained in step S1, a bilateral matching scheme is constructed using a preference ranking mechanism and the Borda counting method to screen out an initial set of abnormal matching pairs between power generators and electricity retailers. This scheme specifically includes the following steps:

[0067] Preference types are set based on the needs of power generators and power retailers. These preference types include bid preference, transaction volume preference, and cooperation stability preference. Bid preference is defined as: ranking the bids of power generators and power retailers, with the highest bidder given priority. Transaction volume preference is defined as: ranking the bids of power generators and power retailers, with the bidder with the most recent bid given priority. Cooperation stability preference is defined as: ranking the bids of power generators and power retailers based on the number of times they have participated in transactions, with the bidder with the most transactions given priority.

[0068] Based on the Borda counting method, the acquired data information is converted into Borda scores:

[0069] For any preferred type of power generator A i To e-commerce seller B j , use the following formula to calculate A i and B j When matching A i Satisfaction

[0070]

[0071] Where n is the total number of electricity sellers; j k is the preference order of electricity sellers for electricity generators;

[0072] Setting: When the generator’s preference order for s retailers is j k When A i and B j When matching A i Satisfaction for

[0073] Then we can get the Borda score matrix S′ under the preference ranking of power generators to electricity retailers: A for

[0074] For any preferred type of seller B j For power generator A i , use the following formula to calculate A i and B j Matching B j Satisfaction

[0075]

[0076] Where m is the total number of power generators; j l is the preference order of generators for retailers;

[0077] Assume that the order of preference of the seller to the s distributors is j. l When A i and B j Matching B j Satisfaction for

[0078] Then we can get the Borda score matrix S′ under the preference ranking of electricity sellers to power generators B for

[0079] Calculate the Borda score matrix S′ under the quotation preference respectively price-A and S′ price-B , Borda score matrix S′ under volume preference volume-A and S′ volume-B And the Borda score matrix S′ under cooperative stability preference stable-A and S′ stable-B ;

[0080] For each Borda score matrix, calculate the corresponding extreme quantile Q 90 (i.e. the top 10% of the matching score), set the satisfaction level to be greater than the extreme quantile Q 90 When is marked as abnormal, the corresponding Boolean matrix is ​​obtained:

[0081]

[0082]

[0083] In the formula is the Boolean matrix of price preference on the power generation side; S′ price-A The element in row i and column j of ; Based on S′ price-A Calculated Q 90 extreme quantiles; is the Boolean matrix of price preference on the electricity sales side; S′ price-B The element in row i and column j of ; Based on S′ price-B Calculated Q 90 extreme quantiles; is the Boolean matrix of transaction volume preference on the generation side; S′ volume-A The element in row i and column j of ; Based on S′ volume-A Calculated Q 90 extreme quantiles; is the Boolean matrix of transaction volume preference on the electricity sales side; S′ volume-B The element in row i and column j of ; Based on S′ volume-B Calculated Q 90 extreme quantiles; is the Boolean matrix of cooperation stability preference on the generation side; S′ stable-A The element in row i and column j of ; Based on S′ stable-A Calculated Q 90 extreme quantiles; is the Boolean matrix of cooperation stability preference for the electricity sales side; S′ stable-B The element in row i and column j of ; Based on S′ stable-B Calculated Q 90 extreme quantiles;

[0084] Will and Merge into the power generation side comprehensive label matrix S A , S A The element in row i and column j The value of

[0085]

[0086] Will and Merge into the comprehensive label matrix S on the power sales side B , S B The element in row i and column j The value of

[0087]

[0088] Comprehensive power generation side comprehensive label matrix S A And the comprehensive label matrix S of the power sales side B , we get the matching matrix S of the two sides of the electricity sales, and the element s in the i-th row and j-th column of S is ij The value of

[0089] According to the element values ​​in the matching matrix S on both sides of the power supply, the matching pairs of power generators and power retailers corresponding to the element value of 1 are screened and used as the initial set of abnormal matching pairs of power generators and power retailers F = {(A i ,B j)|s ij =1}; At this time, the abnormal match is a high-risk element in the initial set;

[0090] S3. Calculate the corresponding joint probability distribution and Shannon entropy value based on the initial set of abnormal matching pairs obtained in step S2; specifically, the following steps are included:

[0091] According to the initial set F of abnormal matching pairs obtained in step S2, for any abnormal matching pair (A i ,B j ), get the generator A i and retailer B j Set the historical transaction price and transaction volume data within the historical time period; set the historical transaction price X and transaction volume Y, and record all possible discrete values ​​of X as x k , all possible discrete values ​​of Y are y l ;

[0092] The following formula is used to calculate the transaction price x k and power l Perform preprocessing:

[0093]

[0094] In the formula is the standardized value of the transaction price; μ x is the mean of the transaction price; σ x is the standard deviation of the transaction price; is the standardized value of transaction volume; μ y is the mean value of transaction volume; σ y is the standard deviation of transaction volume;

[0095] The joint probability distribution f(x,y) of X and Y is calculated using the following formula:

[0096]

[0097] Where M is the number of power generators; N is the number of electricity sellers; h x is the bandwidth parameter of the Gaussian kernel function of the transaction price; h y is the bandwidth parameter of the Gaussian kernel function of the transaction volume; K is the Gaussian kernel function;

[0098] The following formula is used to calculate the power supplier A i and retailer B j The joint entropy of the transaction price and transaction volume H(X,Y):

[0099]

[0100] The Shannon entropy H(X) of the transaction price X is calculated using the following formula:

[0101]

[0102] Where P(x k ) is the probability distribution of transaction price;

[0103] The Shannon entropy H(Y) of the transaction volume Y is calculated using the following formula:

[0104]

[0105] In the formula Q(y l ) is the probability distribution of transaction volume;

[0106] S4. Calculate the joint entropy and mutual information based on the Shannon entropy value obtained in step S3; specifically comprising the following steps:

[0107] According to the obtained joint entropy H(X,Y), the generator A is calculated i and retailer B j The mutual information between the transaction price and the transaction volume I(X;Y) is I(X;Y)=H(X)+H(Y)-H(X,Y);

[0108] Among them, the greater the mutual information, the stronger the behavioral coordination, which is unacceptable in the power market;

[0109] S5. Based on the calculation results obtained in step S4, screen high-risk trading pairs to identify market power risks on both sides of the electricity market. Specifically, the steps include:

[0110] Set the mutual information threshold; in specific implementation, the expert scoring method can be used to obtain the information threshold

[0111] If the power generation side A obtained in step S4 i and electricity sales side B j If the mutual information I(X; Y) between them is greater than the mutual information threshold, it is determined that the power generation side A i and electricity sales side B j There are high-risk trading pairs between them;

[0112] In addition, differentiated settings can be made for different regions or time periods based on actual conditions: for example, the mutual information threshold can be appropriately lowered during periods of tight supply and demand, while it can be increased during normal periods. At the same time, for high-frequency trading pairs or affiliated companies (such as the power generation and sales sides under the same parent company), the mutual information threshold can be adaptively adjusted.

[0113] like Figure 2The figure shows a schematic diagram of the functional modules of the system of the present invention: the system disclosed by the present invention for realizing the method for identifying market power risks on both sides of the electricity market comprises a data acquisition module, an initial matching module, an entropy calculation module, a mutual information calculation module and a risk identification module; the data acquisition module, the initial matching module, the entropy calculation module, the mutual information calculation module and the risk identification module are connected in series in sequence; the data acquisition module is used to obtain transaction data information on the power generation side and the power sales side of the target power market, and upload the data information to the initial matching module; the initial matching module is used to construct a bilateral data structure based on the Borda counting method according to the received data information and the acquired data information through a preference sorting mechanism. Matching plan, screen out the initial set of abnormal matching pairs of power generators and power sellers, and upload the data information to the entropy calculation module; the entropy calculation module is used to calculate the corresponding joint probability distribution and Shannon entropy value based on the received data information and the obtained initial set of abnormal matching pairs, and upload the data information to the mutual information calculation module; the mutual information calculation module is used to calculate the joint entropy and mutual information based on the received data information and the obtained Shannon entropy value, and upload the data information to the risk identification module; the risk identification module is used to screen high-risk transaction pairs based on the received data information and the obtained calculation results, so as to complete the identification of market power risks on both sides of the electricity market.

[0114] In the specific implementation, the system is realized based on the SpringCloud microservice architecture.

[0115] The solution of the present invention aims at the problem of vertical collusion that may exist between the power generation side and the power sales side in the power market. It integrates the bilateral matching model based on Borda number and the entropy balance matching scheme to form a two-layer identification mechanism of "behavior initial screening + result verification".

[0116] The solution of the present invention adopts the Borda counting method, which can rank multiple transaction feature dimensions between power generators and power retailers, such as quotation level, transaction volume matching, historical cooperation frequency, etc., and assign different weights, so as to comprehensively assess the risk level of each transaction pair and effectively capture the active collaboration tendency among market players; the matching matrix generated by the Borda counting method can be used as a preliminary screening tool to help regulators narrow the scope of investigation and improve efficiency.

[0117] The solution of the present invention verifies the rationality of bilateral matching results from a clearing perspective and reflects the risks of market manipulation or collusion in the market, thereby ensuring that the matching results are fair and diverse in multiple dimensions. It can also improve the accuracy of identifying abnormal cooperation or collusion between power generators and electricity sellers.

[0118] The scheme introduces joint entropy and mutual information in information theory, and constructs a data-driven cooperative behavior detection framework; the framework can quantify the nonlinear correlation between the generation side and the electricity sales side transaction behavior, and screen out abnormal pairs with high matching degree and strong cooperation, and finally distinguish between legal cooperation (such as long-term contracts) and illegal collusion, solve the problem that the generation and sales sides cannot identify the cooperative transaction, and improve the accuracy of risk identification.

[0119] The application adopts subjective preference analysis of Borda counting method and objective distribution analysis of entropy balance; the complementary design significantly improves the robustness of the model, can resist strategic interference at the behavior layer, and can penetrate complex noise at the result layer, and provides protection for dynamic risk identification of the power market.

Claims

1. A method for identifying market power risks on both sides of the electricity market, comprising the following steps: S1. Obtain transaction data information on the power generation and sales sides of the target power market; S2. Based on the data information obtained in step S1, a bilateral matching scheme is constructed using a preference ranking mechanism and the Borda counting method to screen out an initial set of abnormal matching pairs between power generators and electricity retailers. S3. Calculate the corresponding joint probability distribution and Shannon entropy value based on the initial set of abnormal matching pairs obtained in step S2; S4. Calculate the joint entropy and mutual information based on the Shannon entropy obtained in step S3; S5. Based on the calculation results obtained in step S4, high-risk transaction pairs are screened to complete the identification of market power risks on both sides of the electricity market.

2. The method for identifying market power risks on both sides of the electricity market according to claim 1 is characterized in that Step S2, based on the data information obtained in step S1, constructs a bilateral matching scheme based on the Borda counting method through a preference sorting mechanism, and screens out an initial set of abnormal matching pairs of power generators and power retailers. Specifically, the following steps are included: Preference types are set based on the needs of power generators and power retailers. These preference types include bid preference, transaction volume preference, and cooperation stability preference. Bid preference is defined as: ranking the bids of power generators and power retailers, with the highest bidder given priority. Transaction volume preference is defined as: ranking the bids of power generators and power retailers, with the bidder with the most recent bid given priority. Cooperation stability preference is defined as: ranking the bids of power generators and power retailers based on the number of times they have participated in transactions, with the bidder with the most transactions given priority. Based on the Borda counting method, the acquired data information is converted into Borda scores: For any preferred type of power generator A i To e-commerce seller B j , use the following formula to calculate A i and B j When matching A i Satisfaction s′ ij A : s′ ij A =n-j k +1 Where n is the total number of electricity sellers; j k is the preference order of electricity sellers for electricity generators; Setting: When the generator’s preference order for s retailers is j k When A i and B j When matching A i Satisfaction s′ ij A s′ ij A =nj k +1-(s-1) / 2; Then we can get the Borda score matrix S′ under the preference ranking of power generators to electricity retailers: A S′ A =[s′ ij A ]; For any preferred type of seller B j For power generator A i , use the following formula to calculate A i and B j Matching B j Satisfaction s′ ji B : s′ ji B =m-j l +1 Where m is the total number of power generators; j l is the preference order of generators for retailers; Assume that the order of preference of the seller to the s distributors is j. l When A i and B j Matching B j Satisfaction s′ ji B s′ ji B =mj l +1-(s-1) / 2; Then we can get the Borda score matrix S′ under the preference ranking of electricity sellers to power generators B S′ B =[s′ ji B ]; Calculate the Borda score matrix S′ under the quotation preference respectively price-A and S′ price-B , Borda score matrix S′ under volume preference volume-A and S′ volume-B And the Borda score matrix S′ under cooperative stability preference stable-A and S′ stable-B ; For each Borda score matrix, calculate the corresponding extreme quantile Q 90 , set when the satisfaction is greater than the extreme quantile Q 90 When is marked as abnormal, the corresponding Boolean matrix is ​​obtained: In the formula is the Boolean matrix of price preference on the power generation side; S′ price-A The element in row i and column j of ; Based on S′ price-A Calculated Q 90 extreme quantiles; is the Boolean matrix of price preference on the electricity sales side; S′ price-B The element in row i and column j of ; Based on S′ price-B Calculated Q 90 extreme quantiles; is the Boolean matrix of transaction volume preference on the generation side; S′ volume-A The element in row i and column j of ; Based on S′ volume-A Calculated Q 90 extreme quantiles; is the Boolean matrix of transaction volume preference on the electricity sales side; S′ volume-B The element in row i and column j of ; Based on S′ volume-B Calculated Q 90 extreme quantiles; is the Boolean matrix of cooperation stability preference on the generation side; S′ stable-A The element in row i and column j of ; Based on S′ stable-A Calculated Q 90 extreme quantiles; is the Boolean matrix of cooperation stability preference for the electricity sales side; S′ stable-B The element in row i and column j of ; Based on S′ stable-B Calculated Q 90 extreme quantiles; Will and Merge into the power generation side comprehensive label matrix S A , S A The element in row i and column j The value of Will and Merge into the comprehensive label matrix S on the power sales side B , S B The element in row i and column j The value of Comprehensive power generation side comprehensive label matrix S A And the comprehensive label matrix S of the power sales side B , we get the matching matrix S of the two sides of the electricity sales, and the element s in the i-th row and j-th column of S is ij The value of According to the element values ​​in the matching matrix S on both sides of the power supply, the matching pairs of power generators and power retailers corresponding to the element value of 1 are screened and used as the initial set of abnormal matching pairs of power generators and power retailers F = {(A i ,B j )|s ij =1}.

3. The method for identifying market power risks on both sides of the electricity market according to claim 2 is characterized in that Step S3, based on the initial set of abnormal matching pairs obtained in step S2, calculates the corresponding joint probability distribution and Shannon entropy value, which specifically includes the following steps: According to the initial set F of abnormal matching pairs obtained in step S2, for any abnormal matching pair (A i ,B j ), get the generator A i and retailer B j Set the historical transaction price and transaction volume data within the historical time period; set the historical transaction price X and transaction volume Y, and record all possible discrete values ​​of X as x k , all possible discrete values ​​of Y are y l ; The following formula is used to calculate the transaction price x k and power l Perform preprocessing: In the formula is the standardized value of the transaction price; μ x is the mean of the transaction price; σ x is the standard deviation of the transaction price; is the standardized value of transaction volume; μ y is the mean value of transaction volume; σ y is the standard deviation of transaction volume; The joint probability distribution f(x,y) of X and Y is calculated using the following formula: Where M is the number of power generators; N is the number of electricity sellers; h x is the bandwidth parameter of the Gaussian kernel function of the transaction price; h y is the bandwidth parameter of the Gaussian kernel function of the transaction volume; K is the Gaussian kernel function; The following formula is used to calculate the power supplier A i and retailer B j The joint entropy of the transaction price and transaction volume H(X,Y): The Shannon entropy H(X) of the transaction price X is calculated using the following formula: Where P(x k ) is the probability distribution of transaction price; The Shannon entropy H(Y) of the transaction volume Y is calculated using the following formula: In the formula Q(y l ) is the probability distribution of transaction volume.

4. The method for identifying market power risks on both sides of the electricity market according to claim 3 is characterized in that The calculation of the joint entropy and mutual information according to the Shannon entropy value obtained in step S3 in step S4 specifically includes the following steps: According to the obtained joint entropy H(X,Y), the generator A is calculated i and retailer B j The mutual information I(X;Y) between the transaction price and the transaction volume is I(X;Y)=H(X)+H(Y)-H(X,Y).

5. The method for identifying market power risks on both sides of the electricity market according to claim 4 is characterized in that Step S5, based on the calculation results obtained in step S4, screens high-risk trading pairs, specifically including the following steps: Set the mutual information threshold; If the power generation side A obtained in step S4 i and electricity sales side B j If the mutual information I(X; Y) between them is greater than the mutual information threshold, it is determined that the power generation side A i and electricity sales side B j There are high-risk trading pairs.

6. A system for implementing the method for identifying market power risks on both sides of the electricity market as claimed in any one of claims 1 to 5, characterized in that It includes a data acquisition module, an initial matching module, an entropy calculation module, a mutual information calculation module and a risk identification module; the data acquisition module, the initial matching module, the entropy calculation module, the mutual information calculation module and the risk identification module are connected in series in sequence; the data acquisition module is used to obtain transaction data information on the power generation side and the power sales side of the target power market, and upload the data information to the initial matching module; The initial matching module is used to construct a bilateral matching scheme based on the Borda counting method according to the received data information through a preference sorting mechanism, screen out the initial set of abnormal matching pairs of power generators and electricity sellers, and upload the data information to the entropy calculation module; The entropy calculation module is used to calculate the corresponding joint probability distribution and Shannon entropy value based on the received data information and the obtained initial set of abnormal matching pairs, and upload the data information to the mutual information calculation module; The mutual information calculation module is used to calculate the joint entropy and mutual information based on the received data information and the obtained Shannon entropy value, and upload the data information to the risk identification module; The risk identification module is used to screen high-risk transaction pairs based on the received data information and the obtained calculation results, so as to complete the identification of market power risks on both sides of the electricity market.

7. The system according to claim 6, characterized in that The system is implemented based on the SpringCloud microservice architecture.

Images