Power resource evaluation configuration method and system based on data driving

Abstract

The invention relates to the field of power resource allocation, in particular to a power resource evaluation configuration method and system based on data driving, and the method comprises the steps: carrying out the clustering of observation data under various evaluation indexes, and carrying out the type division of power market subjects; calculating the distance between different classes under the same evaluation index, obtaining the discrete degree of each evaluation index based on the distance, and obtaining an objective weight; calculating the pre-constructed hierarchical structure model of the power system to obtain a subjective weight; performing weighted summation on the objective weight and the subjective weight to obtain a comprehensive weight of each evaluation index; and performing weighted summation on each evaluation index by taking the comprehensive weight of each evaluation index as a coefficient to obtain a comprehensive score of the corresponding electricity market subject, and performing electricity resource allocation on the electricity market subject based on the comprehensive score. In the evaluation process, both subjective factors and objective factors are considered, power resource distribution is more reasonable, and the customer experience of power utilization subjects is improved.

Description

Technical Field

[0001] The present invention relates to the field of power resource allocation, and in particular to a data-driven power resource evaluation and allocation method and system. Background Art

[0002] With the development of power equipment, the number of power market entities has increased significantly. The electricity consumption characteristics of power market entities vary greatly, resulting in increased difficulty in power resource allocation. Existing technologies all evaluate the electricity consumption of power market entities based on data statistics or experience and allocate power resources, unable to allocate power resources considering multiple aspects, leading to unreasonable power resource allocation, loss of fairness in power resource allocation, and poor experience for power market entities.

[0003] Chinese Patent CN106058851 provides a power resource allocation method and system based on demand response. Beforehand, it obtains the first response load value with the largest first parameter value according to the predicted value of the influence parameter of the user reference load model, and then allocates power resources. Afterward, it obtains the second response load value with the largest second parameter value according to the actual value of the influence parameter of the user baseline load model, and then adjusts the already allocated power resources. It can only evaluate and allocate based on the objective indicators of electricity users, without considering the subjective influencing factors of electricity users, and the user experience is not good. Summary of the Invention

[0004] To solve the problems of unreasonable power resource allocation and poor user experience in the prior art, in the first aspect of the present invention, a data-driven power resource evaluation and allocation method is proposed, including:

[0005] Clustering the observed data under multiple evaluation indicators to classify power market entities;

[0006] Calculating the distance between different classes under the same evaluation indicator, obtaining the dispersion degree of each evaluation indicator based on the distance, and obtaining the objective weight of the corresponding evaluation indicator by taking the ratio of the dispersion degree of a single evaluation indicator to the sum of the dispersion degrees of all evaluation indicators;

[0007] Calculating the subjective weight of each evaluation indicator by calculating the pre-constructed power system hierarchical structure model;

[0008] Performing weighted summation on the objective weight and the subjective weight to obtain the comprehensive weight of each evaluation indicator;

[0009] Performing weighted summation on each evaluation indicator with the comprehensive weight of each evaluation indicator as the coefficient to obtain the comprehensive score of the corresponding power market entity, and allocating power resources to the power market entity based on the comprehensive score.

[0010] Optionally, the evaluation indicators include:

[0011] The accuracy rate of declared operation parameters, the completion rate of power quantity compliance, the completion rate of ancillary services, the reliable rate of equipment operation, or the reliable rate of coal supply for electricity.

[0012] Optionally, the calculation formula for the accuracy rate of declared operation parameters is:

[0013]

[0014] Where, is the accuracy rate of declared operation parameters of the i-th electricity market entity, r i c is the number of incorrectly declared parameters of the i-th electricity market entity, r i y is the number of omitted declared parameters of the i-th electricity market entity, r i is the total number of declared parameters of the i-th electricity market entity.

[0015] Optionally, the calculation formula for the completion rate of power quantity compliance is:

[0016]

[0017] Where, is the completion rate of power quantity compliance of the i-th electricity market entity, is the actual settlement power quantity of the i-th electricity market entity, is the total winning bid power quantity of the i-th electricity market entity.

[0018] Optionally, the calculation formula for the completion rate of ancillary services is:

[0019]

[0020] Where, is the completion rate of ancillary services of the i-th electricity market entity, is the actual power of the i-th electricity market entity, is the declared power of the i-th electricity market entity.

[0021] Optionally, the calculation formula for the reliable rate of equipment operation is:

[0022]

[0023] Where, is the reliable rate of equipment operation of the i-th electricity market entity, is the planned operation time of the i-th electricity market entity, is the unplanned outage time of the i-th electricity market entity.

[0024] Optionally, the calculation formula for the reliable rate of thermal coal supply is as follows:

[0025]

[0026] Wherein, is the reliable rate of thermal coal supply for the i-th electricity market entity, is the actual thermal coal supply volume of the i-th electricity market entity, is the planned thermal coal supply volume of the i-th electricity market entity.

[0027] Optionally, the clustering of the observed data under multiple evaluation indicators to classify electricity market entities specifically includes:

[0028] Using the Canopy clustering algorithm to conduct rough classification on the observed data under multiple evaluation indicators, and determining the number of clusters K and the initial cluster centers;

[0029] Taking the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of electricity market entities using the K-means clustering algorithm.

[0030] Optionally, the steps of using the Canopy clustering algorithm to conduct rough classification on the observed data under multiple evaluation indicators and determining the number of clusters K and the initial cluster centers include:

[0031] S111: Set initial distance thresholds T1 and T2 based on the observed data, where T1 > T2;

[0032] S112: Arbitrarily select an electricity market entity P, and calculate the nearest distance D between the electricity market entity P and all cluster centers respectively;

[0033] S113: If the nearest distance D is less than or equal to T2, then eliminate the electricity market entity P; if the nearest distance D is less than T1 and greater than T2, then add the electricity market entity P to the cluster where the nearest cluster center is located; if the nearest distance D is greater than or equal to T1, then the electricity market entity P forms a new cluster;

[0034] S114: Repeat all of the above S112 and S113 until all electricity market entities are classified.

[0035] Optionally, the steps of taking the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of electricity market entities using the K-means clustering algorithm include:

[0036] S121: Calculate the distances from each electricity market entity to the K clustering centers, and divide each electricity market entity into the class corresponding to the clustering center with the minimum distance;

[0037] S122: For each class, recalculate its clustering center;

[0038] S123: Repeat S121 and S122 until a preset termination condition is reached to obtain the final type division of the electricity market entities.

[0039] Optionally, the calculation formula for the clustering center in S122 is:

[0040]

[0041] where a j is the new clustering center, |c i | is the number of electricity market entities in the c i th class, and t is the value of the electricity market entity belonging to this class.

[0042] Optionally, calculating the distances between different classes under the same evaluation index and obtaining the dispersion degree of each evaluation index based on the distances specifically includes:

[0043] Calculate the distances between any two clustering centers of each evaluation index;

[0044] Use the variance of all distances under the same evaluation index as the dispersion degree of the evaluation index.

[0045] Optionally, the calculation formula for the dispersion degree is:

[0046]

[0047] where σ m is the dispersion degree of the mth evaluation index, K m is the final number of clusters of the observed data of the mth evaluation index, is the distance between the rth clustering center and the jth clustering center in the mth evaluation index, is the average value of the distances between the clustering centers under the mth evaluation index.

[0048] Optionally, the power system hierarchical structure model includes an objective layer, a criterion layer, and an index layer, where:

[0049] The objective layer is used to define the evaluation objective when allocating power resources;

[0050] The criterion layer is used to determine the main factors affecting the evaluation objective;

[0051] The index layer is used to determine the multiple evaluation indexes.

[0052] Optionally, calculating the subjective weights of the evaluation indicators from the pre-constructed power system hierarchical model includes the steps of:

[0053] Comparing each element in the criterion layer pairwise to obtain the criterion layer judgment matrix; and comparing each element in the index layer pairwise to obtain the index layer judgment matrix;

[0054] Conducting a consistency test on the criterion layer judgment matrix and assigning weights to obtain the weights of each element in the criterion layer, and conducting a consistency test on the index layer judgment matrix and assigning weights to obtain the weights of each element in the index layer;

[0055] Multiplying the weights of each evaluation indicator relative to different indicators by the weights of the elements in the criterion layer and summing them to obtain the subjective weights of each evaluation indicator.

[0056] Optionally, conducting a consistency test on the criterion layer judgment matrix and assigning weights to obtain the weights of each element in the criterion layer, and conducting a consistency test on the index layer judgment matrix and assigning weights to obtain the weights of each element in the index layer includes:

[0057] Conducting a consistency test on the criterion layer judgment matrix. If the test is passed, the eigenvalue method, geometric mean method, or arithmetic mean method is used to calculate the weights of each element in the criterion layer relative to the target layer;

[0058] Conducting a consistency test on the index layer judgment matrix. If the test is passed, the eigenvalue method, geometric mean method, or arithmetic mean method is used to calculate the weights of each element in the index layer relative to the criterion layer.

[0059] Optionally, the calculation formula for the comprehensive score of the power market entity is:

[0060]

[0061] where s i is the comprehensive score of the i-th power market entity, is the reporting accuracy rate of the operation parameters of the i-th power market entity, is the electricity quantity performance completion rate of the i-th power market entity, is the auxiliary service completion rate of the i-th power market entity, is the equipment operation reliability rate of the i-th power market entity, is the reliable rate of coal supply for electricity of the i-th power market entity, w1 is the weight coefficient of the reporting accuracy rate of operation parameters, w2 is the weight coefficient of the electricity quantity performance completion rate, w3 is the weight coefficient of the auxiliary service completion rate, w4 is the weight coefficient of the equipment operation reliability rate, and w5 is the weight coefficient of the reliable rate of coal supply for electricity.

[0062] Optionally, the calculation formula for the objective weight is as follows:

[0063]

[0064] Wherein, is the objective weight of the m-th evaluation index, and σ r is the degree of dispersion of the r-th evaluation index, N is the total number of evaluation indexes, and σ m is the degree of dispersion of the m-th evaluation index.

[0065] Optionally, the observed data is obtained from various data sources by using data mining and / or web crawling technologies.

[0066] The second aspect of the present invention provides a data-driven power resource evaluation and allocation system, including:

[0067] Partition module: used for clustering the observed data under multiple evaluation indexes to classify power market entities;

[0068] Objective weight acquisition module: used for calculating the distance between different classes under the same evaluation index, obtaining the degree of dispersion of each evaluation index based on the distance, and obtaining the objective weight of the corresponding evaluation index by taking the ratio of the degree of dispersion of a single evaluation index to the sum of the degrees of dispersion of all evaluation indexes;

[0069] Subjective weight acquisition module: used for calculating the subjective weights of each evaluation index by calculating a pre-constructed power system hierarchical structure model;

[0070] Comprehensive weight acquisition module: used for weighted summing the objective weight and the subjective weight to obtain the comprehensive weight of each evaluation index;

[0071] Comprehensive score acquisition module: used for weighted summing each evaluation index with the comprehensive weight of each evaluation index as a coefficient to obtain the comprehensive score of the corresponding power market entity, and allocating power resources to the power market entity based on the comprehensive score.

[0072] Optionally, the evaluation indexes in the partition module include:

[0073] Accuracy rate of operation parameter declaration, completion rate of power quantity performance, completion rate of auxiliary service, equipment operation reliability rate, or reliable rate of coal supply for power generation.

[0074] Optionally, the calculation formula for the accuracy rate of operation parameter declaration in the partition module is as follows:

[0075]

[0076] Wherein, is the accuracy rate of operation parameter declaration of the i-th power market entity, and ri c is the number of incorrectly declared parameters of the i-th electricity market entity, r i y is the number of omitted declared parameters of the i-th electricity market entity, r i is the total number of declared parameters of the i-th electricity market entity.

[0077] Optionally, the calculation formula for the electricity quantity compliance completion rate in the division module is:

[0078]

[0079] where is the electricity quantity compliance completion rate of the i-th electricity market entity, is the actual settlement electricity quantity of the i-th electricity market entity, is the total winning bid electricity quantity of the i-th electricity market entity.

[0080] Optionally, the calculation formula for the ancillary service completion rate in the division module is:

[0081]

[0082] where is the ancillary service completion rate of the i-th electricity market entity, is the actual power of the i-th electricity market entity, is the declared power of the i-th electricity market entity.

[0083] Optionally, the calculation formula for the equipment operation reliability rate in the division module is:

[0084]

[0085] where is the equipment operation reliability rate of the i-th electricity market entity, is the planned operation time of the i-th electricity market entity, is the unplanned outage time of the i-th electricity market entity.

[0086] Optionally, the calculation formula for the steam coal supply reliability rate in the division module is:

[0087]

[0088] where is the steam coal supply reliability rate of the i-th electricity market entity, is the actual steam coal supply quantity of the i-th electricity market entity, is the planned steam coal supply quantity of the i-th electricity market entity.

[0089] Optionally, the partitioning module clusters the observed data under multiple evaluation metrics to classify the electricity market entities, specifically including:

[0090] Coarsely classify the observed data under multiple evaluation metrics using the Canopy clustering algorithm to determine the number of clusters K and the initial cluster centers;

[0091] Use the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of the electricity market entities using the K-means clustering algorithm.

[0092] Optionally, the partitioning module coarsely classifies the observed data under multiple evaluation metrics using the Canopy clustering algorithm to determine the number of clusters K and the initial cluster centers. The steps include:

[0093] S111: Set initial distance thresholds T1 and T2 based on the observed data, where T1 > T2;

[0094] S112: Arbitrarily select an electricity market entity P and calculate the closest distance D between the electricity market entity P and all cluster centers respectively;

[0095] S113: If the closest distance D is less than or equal to T2, then eliminate the electricity market entity P; if the closest distance D is less than T1 and greater than T2, then add the electricity market entity P to the cluster corresponding to the closest cluster center; if the closest distance D is greater than or equal to T1, then the electricity market entity P forms a new cluster;

[0096] S114: Repeat all the above steps 112 and 113 until all the electricity market entities are classified.

[0097] Optionally, the partitioning module uses the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of the electricity market entities using the K-means clustering algorithm. The steps include:

[0098] S121: Calculate the distances from each electricity market entity to the K cluster centers and divide each electricity market entity into the class corresponding to the cluster center with the minimum distance;

[0099] S122: For each class, recalculate its cluster center;

[0100] S123: Repeat S121 and S122 until a preset termination condition is reached to obtain the final classification of the electricity market entities.

[0101] Optionally, the calculation formula for the cluster center in the partitioning module is:

[0102]

[0103] Among them, a j is the new clustering center, |c i | is the number of power market entities in the c i th class, and t is the value of the power market entities belonging to this class.

[0104] Optionally, the objective weight acquisition module calculates the distance between different classes under the same evaluation index, and obtains the dispersion degree of each evaluation index based on the distance, specifically including:

[0105] Calculate the distance between any two clustering centers of each evaluation index;

[0106] Use the variance of all distances under the same evaluation index as the dispersion degree of the evaluation index.

[0107] Optionally, the calculation formula for the dispersion degree in the objective weight acquisition module is:

[0108]

[0109] Among them, σ m is the dispersion degree of the mth evaluation index, K m is the final number of clusters of the observed data of the mth evaluation index, is the distance between the rth clustering center and the jth clustering center in the mth evaluation index, is the average value of the distances between the clustering centers under the mth evaluation index.

[0110] Optionally, the power system hierarchical structure model in the subjective weight acquisition module includes an objective layer, a criterion layer, and an index layer, where:

[0111] The objective layer is used to define the evaluation objective when allocating power resources;

[0112] The criterion layer is used to determine the main factors affecting the evaluation objective;

[0113] The index layer is used to determine the multiple evaluation indexes.

[0114] Optionally, the subjective weight acquisition module calculates the subjective weights of each evaluation index by calculating the pre-constructed power system hierarchical structure model. The steps include:

[0115] Compare each element in the criterion layer pairwise to obtain the criterion layer judgment matrix; and compare each element in the index layer pairwise to obtain the index layer judgment matrix;

[0116] Perform a consistency test on the criterion layer judgment matrix and assign weights to obtain the weights of each element in the criterion layer. Perform a consistency test on the index layer judgment matrix and assign weights to obtain the weights of each element in the index layer;

[0117] Multiply the weights of each evaluation index relative to different indexes by the weights of the elements in the criterion layer and sum them to obtain the subjective weights of each evaluation index.

[0118] Optionally, the subjective weight acquisition module performs a consistency test on the criterion layer judgment matrix and assigns weights to obtain the weights of each element in the criterion layer. Performing a consistency test on the index layer judgment matrix and assigning weights to obtain the weights of each element in the index layer includes:

[0119] Perform a consistency test on the criterion layer judgment matrix. If the test passes, use the eigenvalue method, geometric mean method, or arithmetic mean method to calculate the weights of each element in the criterion layer relative to the target layer;

[0120] Perform a consistency test on the index layer judgment matrix. If the test passes, use the eigenvalue method, geometric mean method, or arithmetic mean method to calculate the weights of each element in the index layer relative to the criterion layer.

[0121] Optionally, the calculation formula for the comprehensive score of the electricity market entity in the comprehensive score acquisition module is:

[0122]

[0123] where s i is the comprehensive score of the i-th electricity market entity, is the reporting accuracy rate of the operation parameters of the i-th electricity market entity, is the electricity quantity performance fulfillment rate of the i-th electricity market entity, is the auxiliary service fulfillment rate of the i-th electricity market entity, is the equipment operation reliability rate of the i-th electricity market entity, is the reliable rate of thermal coal supply of the i-th electricity market entity, w1 is the weight coefficient of the reporting accuracy rate of operation parameters, w2 is the weight coefficient of the electricity quantity performance fulfillment rate, w3 is the weight coefficient of the auxiliary service fulfillment rate, w4 is the weight coefficient of the equipment operation reliability rate, and w5 is the weight coefficient of the reliable rate of thermal coal supply.

[0124] Optionally, the calculation formula for the objective weight in the objective weight acquisition module is:

[0125]

[0126] where, is the objective weight of the m-th evaluation index, σ ris the degree of dispersion of the r-th evaluation index, N is the total number of evaluation indexes, and σ m is the degree of dispersion of the m-th evaluation index.

[0127] Optionally, the observed data in the partitioning module is obtained from various data sources by using data mining and / or web crawling technology.

[0128] In a third aspect, the present invention also provides a computing device, including: at least one processor and a memory;

[0129] The memory is used to store one or more programs;

[0130] When the one or more programs are executed by the one or more processors, a data-driven power resource evaluation and configuration method as described above is implemented.

[0131] In a fourth aspect, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, a data-driven power resource evaluation and configuration method as described above is implemented.

[0132] Compared with the prior art, the beneficial effects of the present invention are:

[0133] The present invention provides a data-driven power resource evaluation and configuration method and system, including clustering observed data under multiple evaluation indexes to classify power market entities; calculating the distance between different classes under the same evaluation index, obtaining the degree of dispersion of each evaluation index based on the distance, obtaining the objective weight of the corresponding evaluation index by taking the ratio of the degree of dispersion of a single evaluation index to the sum of the degrees of dispersion of all evaluation indexes; calculating the subjective weight of each evaluation index for a pre-constructed power system hierarchical structure model; performing weighted summation on the objective weight and the subjective weight to obtain the comprehensive weight of each evaluation index; performing weighted summation on each evaluation index with the comprehensive weight of each evaluation index as a coefficient to obtain the comprehensive score of the corresponding power market entity, and allocating power resources to the power market entity based on the comprehensive score; in the present invention, subjective and objective joint evaluation is adopted under multiple evaluation indexes to obtain a comprehensive weight coefficient, and the evaluation index is corrected by using the comprehensive weight coefficient. Therefore, both subjective factors and objective factors are considered in the evaluation process, and the power resource allocation is more reasonable, which can improve the customer experience of electricity users. BRIEF DESCRIPTION OF THE DRAWINGS

[0134] Figure 1 is a flowchart of the data-driven power resource evaluation and configuration method proposed by the present invention;

[0135] Figure 2 proposed by the present invention Figure 1Schematic diagram of the detailed steps of step S1 in

[0136] Figure 3 Proposed by the present invention Figure 2 Schematic diagram of the detailed steps of step S11 in

[0137] Figure 4 Proposed by the present invention Figure 2 Schematic diagram of the detailed steps of step S12 in

[0138] Figure 5 Proposed by the present invention Figure 1 Schematic diagram of the detailed steps of step S2 in

[0139] Figure 6 Proposed by the present invention Figure 1 Schematic diagram of the detailed steps of step S3 in

[0140] Figure 7 Schematic diagram of the structure of the data-driven power resource evaluation and configuration system proposed by the present invention

[0141] Figure 8 Schematic diagram of the structure of the electronic device proposed by the present invention. Detailed implementation manners

[0142] The present invention proposes a data-driven method, system, device and medium for evaluating and allocating electric power resources. First, clustering algorithms are used to divide power market entities into different categories under each evaluation index. Obviously, power market entities with similar characteristics will be grouped into one category. Under a certain evaluation index, if the classification of power market entities is very concentrated, it means that under this index, power market entities have great similarity, and the evaluation results obtained based on this evaluation index have no discrimination, that is, the corresponding index measurement standard is weak. On the contrary, if under a certain evaluation index, the classification of power market entities is very dispersed, it means that under this evaluation index, power market entities have great distinctiveness, and the evaluation results obtained based on this evaluation index have discrimination, that is, the corresponding evaluation index measurement standard is strong. Then, the ratio of the dispersion degree of a single evaluation index to the sum of the dispersion degrees of all evaluation indexes is used as the objective weight of the corresponding evaluation index; then, the power system hierarchical structure model is adopted to identify the importance degree of each evaluation index relative to evaluating power market entities through the analytic hierarchy process. Obviously, the indexes that have a significant impact on the safe operation of the power system and market order should receive special attention, and higher weights should be assigned to these key evaluation indexes accordingly. The present invention uses the linear combination of objective weight and subjective weight as the coefficient of the linear equation to obtain the comprehensive score of power market entities. The weight distribution combining subjectivity and objectivity eliminates the information loss caused by index elimination in traditional methods, and at the same time considers the influence of the subjective behavior of power market entities on the allocation of electric power resources, so that the allocation of electric power resources can be more reasonable. The solutions in the present invention will be introduced in detail below with reference to the accompanying drawings.

[0143] Embodiment 1:

[0144] A data-driven method for evaluating and allocating electric power resources, as Figure 1 shown, includes the following steps S1 to S5:

[0145] S1: Cluster the observed data under multiple evaluation indexes to classify power market entities.

[0146] In a further preferred solution, the evaluation indexes include:

[0147] The accuracy rate of declared operation parameters, the completion rate of power quantity performance, the completion rate of auxiliary services, the reliable rate of equipment operation or the reliable rate of coal supply for electricity. In the embodiment of the present invention, these five indexes are used to evaluate each power market entity.

[0148] In a further preferred solution, the calculation formula of the accuracy rate of declared operation parameters is:

[0149]

[0150] Where, is the declaration accuracy rate of the operating parameters of the i-th electricity market entity, r i c is the number of wrongly declared parameters of the i-th electricity market entity, r i y is the number of omitted declared parameters of the i-th electricity market entity, r i is the total number of declared parameters of the i-th electricity market entity.

[0151] In a further optimized solution, the calculation formula for the electricity quantity compliance completion rate is:

[0152]

[0153] where is the electricity quantity compliance completion rate of the i-th electricity market entity, is the actual settlement electricity quantity of the i-th electricity market entity, is the total winning bid electricity quantity of the i-th electricity market entity.

[0154] In a further optimized solution, the calculation formula for the ancillary service completion rate is:

[0155]

[0156] where is the ancillary service completion rate of the i-th electricity market entity, is the actual power of the i-th electricity market entity, is the declared power of the i-th electricity market entity.

[0157] In a further optimized solution, the calculation formula for the equipment operation reliability rate is:

[0158]

[0159] where is the equipment operation reliability rate of the i-th electricity market entity, is the planned operation time of the i-th electricity market entity, is the unplanned outage time of the i-th electricity market entity.

[0160] In a further optimized solution, the calculation formula for the reliable rate of thermal coal supply is:

[0161]

[0162] where is the reliable rate of thermal coal supply of the i-th electricity market entity, is the actual thermal coal supply quantity of the i-th electricity market entity, is the planned supply of thermal coal for the i-th electricity market entity.

[0163] In a further optimized solution, clustering the observed data under multiple evaluation indicators to classify electricity market entities, such as Figure 2 shown, specifically including steps S11 and S12:

[0164] S11: Coarsely classify the observed data under multiple evaluation indicators using the Canopy clustering algorithm to determine the number of clusters K and the initial cluster centers.

[0165] In a further optimized solution, the observed data can be obtained from various data sources using data mining and / or web crawler technology.

[0166] The collected data also needs to be preprocessed, including cleaning, denoising, duplicate removal, and standardization preprocessing operations. Among them, data cleaning is used to detect and correct errors, missing values, outliers, etc. in the data, which includes removing duplicate data, handling missing values, detecting and handling outliers, etc. Data denoising and duplicate removal are used to delete noise and duplicate items in the data to reduce interference with the model. Data standardization is used to transform the data into a standard form with unified metrics and ranges to ensure consistent data trade-offs for different features. An optional preprocessing operation process is as follows: First, delete duplicate data items by comparing the unique identifiers of the data items; then use linear interpolation to fill in the missing values; and use an outlier detection algorithm to identify outliers and delete the outliers: then use a filter to remove the noise in the data; finally, use the Min-Max standardization method to transform the data into values within the range of [0,1].

[0167] In this step S11, electricity market entities are classified into different classes using a clustering algorithm under each evaluation indicator, and the distances between different classes under each evaluation indicator are calculated. This step measures the evaluation ability of the evaluation indicators by clustering electricity market entities under different indicators spontaneously: The more dispersed the clustering result of electricity market entities under a certain evaluation indicator, the stronger the evaluation ability of this evaluation indicator; the more concentrated the clustering result of electricity market entities under a certain evaluation indicator, the weaker the evaluation ability of this evaluation indicator.

[0168] Canopy clustering is a hierarchical clustering method that can help determine the number of clusters K and the initial cluster centers in K-means clustering. Through Canopy clustering, data points that may belong to the same cluster can be initially screened out, thereby reducing the amount of data that needs to be processed by the subsequent K-means clustering algorithm. And Canopy clustering provides a more stable initial clustering center for K-means clustering by eliminating the influence of noise and outliers, reducing the problem that K-means clustering is sensitive to the initial center. Moreover, Canopy clustering can determine the appropriate number of clusters K according to the actual distribution of the data, rather than subjectively setting the value of K.

[0169] In a further preferred embodiment, as Figure 3 shown, step S11 specifically includes the following steps S111 to S114:

[0170] S111: Set initial distance thresholds T1 and T2 based on the observed data, where T1 > T2;

[0171] S112: Arbitrarily select a power market entity P, and calculate the nearest distance D between the power market entity P and all cluster centers respectively;

[0172] S113: If the nearest distance D is less than or equal to T2, then the power market entity P is removed; if the nearest distance D is less than T1 and greater than T2, then the power market entity P is added to the cluster where the cluster center closest to it is located; if the nearest distance D is greater than or equal to T1, then the power market entity P forms a new cluster;

[0173] S114: Repeat all of the above S112 and S113 until all the power market entities are classified.

[0174] S12: Use the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final type division of the power market entities using the K-means clustering algorithm.

[0175] K-means clustering is a commonly used unsupervised machine learning algorithm for dividing a data set into K clusters, where each cluster consists of data points with relatively high similarity in the data set. The working principle of K-means clustering is to minimize the sum of the distances between each data point and the center point of the cluster to which it belongs through iterative optimization. The concept and implementation of the K-means algorithm are very simple, and the computational complexity of the algorithm is relatively low, making it suitable for processing large-scale data sets. However, K-means clustering also has some limitations, such as being sensitive to the selection of the initial center and being prone to falling into local optimal solutions. Therefore, in this application, the power market entities are first roughly clustered through the Canopy clustering algorithm to determine the number of clusters K and the initial cluster centers of the k-means clustering.

[0176] Take the number of clusters \(K\) and the initial cluster centers as the input of the K-means clustering algorithm, and use the K-means clustering algorithm to divide several electricity market players into different classes. For example Figure 4 As shown, the working steps of the K-means clustering algorithm include steps S121 to S123:

[0177] S121: Calculate the distances from each electricity market player to the \(K\) cluster centers, and divide each electricity market player into the class corresponding to the cluster center with the minimum distance.

[0178] Determine that the number of clusters is \(K\) and the initial cluster centers are \(a = \{a_1, a_2, \ldots, a_K\}\).

[0179] S122: For each class, recalculate its cluster center.

[0180] The calculation formula for the cluster center is:

[0181]

[0182] where \(a\) j is the new cluster center, \(|c|\) i is the number of electricity market players in the \(c\)-th i class, and \(t\) is the value of the electricity market players belonging to this class.

[0183] S123: Repeat S121 and S122 until a preset termination condition is reached, and obtain the final type division of the electricity market players.

[0184] Through the above steps, each electricity market player can be divided into different classes under each evaluation index. After dividing each electricity market player into different classes under a certain evaluation index, each class has a cluster center, which can be obtained by taking the average or the mode of all the values of the class to which it belongs. The distance between any two cluster centers can be used as the distance between the classes.

[0185] S2: Calculate the distances between different classes under the same evaluation index, obtain the dispersion degree of each evaluation index based on the distances, and obtain the objective weight of the corresponding evaluation index by taking the ratio of the dispersion degree of a single evaluation index to the sum of the dispersion degrees of all evaluation indexes.

[0186] In a further optimized solution, the calculation of the distances between different classes under the same evaluation index and obtaining the dispersion degree of each evaluation index based on the distances, as Figure 5 shown, specifically includes the following steps S21 and S22:

[0187] S21: Calculate the distances between any two cluster centers of each evaluation index; the Euclidean distance between the two cluster centers can be used.

[0188] S22: Use the variance of all distances under the same evaluation index as the degree of dispersion of the evaluation index.

[0189] The calculation formula for the degree of dispersion is as follows:

[0190]

[0191] where σ m is the degree of dispersion of the m-th evaluation index, K m is the final number of clusters of the observed data of the m-th evaluation index, is the distance between the r-th cluster center and the j-th cluster center in the m-th evaluation index, is the average value of the distances of the cluster centers under the m-th evaluation index.

[0192] In a further optimized scheme, the calculation formula for the objective weight is as follows:

[0193]

[0194] where, is the objective weight of the m-th evaluation index, σ m is the degree of dispersion of the m-th evaluation index, N is the total number of evaluation indexes, and σ r is the degree of dispersion of the r-th evaluation index.

[0195] S3: Calculate the subjective weights of each evaluation index for the pre-constructed power system hierarchical structure model.

[0196] Construct a power system hierarchical structure model, conduct pairwise comparisons on each layer of elements to establish a judgment matrix, multiply the weights of each index relative to different criteria by the weights of the criterion layer and sum them to obtain the subjective weights of each index.

[0197] In a further optimized scheme, the power system hierarchical structure model includes an objective layer, a criterion layer, and an index layer, where:

[0198] The objective layer is used to define the evaluation objective during power resource allocation; it takes grid operation safety and market order as the evaluation orientation;

[0199] The criterion layer is used to determine the main factors affecting the evaluation objective; the criterion layer determines the main factors affecting grid operation safety and market order, including operation standardization, performance fulfillment ability, emergency response ability, and operation stability;

[0200] The index layer is used to determine the multiple evaluation indexes, including: accuracy rate of operation parameter declaration, completion rate of power quantity fulfillment, completion rate of auxiliary service, equipment operation reliability rate, or reliable rate of thermal coal supply.

[0201] In a further preferred embodiment, calculating the subjective weights of the evaluation indicators by calculating the pre-constructed hierarchical model of the power system, as Figure 6 shown, the steps include steps S31 to S33:

[0202] S31: Compare each element in the criterion layer pairwise to obtain the criterion layer judgment matrix; and compare each element in the index layer pairwise to obtain the index layer judgment matrix;

[0203] S32: Conduct a consistency test on the criterion layer judgment matrix and assign weights to obtain the weights of each element in the criterion layer, and conduct a consistency test on the index layer judgment matrix and assign weights to obtain the weights of each element in the index layer;

[0204] S33: Multiply the weights of each evaluation indicator relative to different indicators by the weights of the elements in the criterion layer and sum them to obtain the subjective weights of each evaluation indicator.

[0205] In a further preferred embodiment, step S32 specifically includes:

[0206] Conduct a consistency test on the criterion layer judgment matrix. If the test is passed, calculate the weights of each element in the criterion layer relative to the target layer using the eigenvalue method, geometric mean method, or arithmetic mean method;

[0207] Conduct a consistency test on the index layer judgment matrix. If the test is passed, calculate the weights of each element in the index layer relative to the criterion layer using the eigenvalue method, geometric mean method, or arithmetic mean method.

[0208] S4: Perform a weighted sum of the objective weights and subjective weights to obtain the comprehensive weights of each evaluation indicator.

[0209] The calculation formula for the comprehensive weight is:

[0210]

[0211] where, w m is the comprehensive weight under the mth evaluation indicator, is the subjective weight under the mth evaluation indicator, is the objective weight under the mth evaluation indicator, and α is the linear combination coefficient, 0 < α < 1.

[0212] Normalize the comprehensive weights of all evaluation indicators to ensure that the weights sum to 1.

[0213] S5: Perform a weighted sum of each evaluation indicator with the comprehensive weight of each evaluation indicator as the coefficient to obtain the comprehensive score of the corresponding power market entity, and allocate power resources to the power market entity based on the comprehensive score.

[0214] Taking the comprehensive weights of each evaluation index as the coefficients of the linear equation, an evaluation model for power market entities is obtained. By inputting the observed data of the power market entities to be evaluated on different evaluation indexes into the evaluation model, the evaluation results are output.

[0215] The calculation formula of the evaluation model for power market entities is as follows:

[0216]

[0217] Where s i is the comprehensive score of the i-th power market entity, is the reporting accuracy rate of the operation parameters declared by the i-th power market entity, is the electricity quantity performance fulfillment rate of the i-th power market entity, is the auxiliary service fulfillment rate of the i-th power market entity, is the equipment operation reliability rate of the i-th power market entity, is the reliable rate of coal supply for electricity of the i-th power market entity, w1 is the weight coefficient of the reporting accuracy rate of operation parameters, w2 is the weight coefficient of the electricity quantity performance fulfillment rate, w3 is the weight coefficient of the auxiliary service fulfillment rate, w4 is the weight coefficient of the equipment operation reliability rate, and w5 is the weight coefficient of the reliable rate of coal supply for electricity.

[0218] In the embodiment of the present invention, the weights of the reporting accuracy rate x y of operation parameters, the electricity quantity performance fulfillment rate x l the auxiliary service fulfillment rate x f the equipment operation reliability rate x s and the reliable rate of coal supply for electricity x g are 44.31%, 17.59%, 28.32%, 7.08%, and 11.69% respectively.

[0219] Based on the above results, subsequently, the evaluation results can be linked to market access, priority trading, auxiliary service fees, and performance guarantee amounts, giving preferential treatment in policies and resources to high-quality entities and taking disciplinary measures against bad entities.

[0220] Embodiment 2:

[0221] Based on the same inventive concept, the present invention also provides a data-driven power resource evaluation and allocation system, as Figure 7 shown, including:

[0222] A division module: used for clustering the observed data under various evaluation indexes to classify power market entities;

[0223] Objective weight acquisition module: used to calculate the distance between different classes under the same evaluation index, obtain the dispersion degree of each evaluation index based on the distance, and obtain the objective weight of the corresponding evaluation index by taking the ratio of the dispersion degree of a single evaluation index to the sum of the dispersion degrees of all evaluation indexes;

[0224] Subjective weight acquisition module: used to calculate the subjective weight of each evaluation index by calculating the pre-constructed power system hierarchical structure model;

[0225] Comprehensive weight acquisition module: used to perform weighted summation on the objective weight and subjective weight to obtain the comprehensive weight of each evaluation index;

[0226] Comprehensive score acquisition module: used to perform weighted summation on each evaluation index with the comprehensive weight of each evaluation index as the coefficient to obtain the comprehensive score of the corresponding power market entity, and allocate power resources to the power market entity based on the comprehensive score.

[0227] In a further preferred solution, the evaluation indexes in the division module include:

[0228] Accuracy rate of operation parameter declaration, completion rate of power quantity compliance, completion rate of auxiliary service, reliable rate of equipment operation, or reliable rate of coal supply.

[0229] In a further preferred solution, the calculation formula for the accuracy rate of operation parameter declaration in the division module is:

[0230]

[0231] Where is the accuracy rate of operation parameter declaration of the i-th power market entity, r i c is the number of wrongly declared parameters of the i-th power market entity, r i y is the number of omitted declared parameters of the i-th power market entity, r i is the total number of declared parameters of the i-th power market entity.

[0232] In a further preferred solution, the calculation formula for the completion rate of power quantity compliance in the division module is:

[0233]

[0234] Where is the completion rate of power quantity compliance of the i-th power market entity, is the actual settled power quantity of the i-th power market entity, is the total winning bid power quantity of the i-th power market entity.

[0235] In a further preferred solution, the calculation formula for the completion rate of auxiliary services in the division module is:

[0236]

[0237] Wherein, is the completion rate of auxiliary services of the i-th electricity market entity, is the actual power of the i-th electricity market entity, is the declared power of the i-th electricity market entity.

[0238] In a further preferred solution, the calculation formula for the equipment operation reliability rate in the division module is:

[0239]

[0240] Wherein, is the equipment operation reliability rate of the i-th electricity market entity, is the planned operation time of the i-th electricity market entity, is the unplanned outage time of the i-th electricity market entity.

[0241] In a further preferred solution, the calculation formula for the reliable rate of thermal coal supply in the division module is:

[0242]

[0243] Wherein, is the reliable rate of thermal coal supply of the i-th electricity market entity, is the actual thermal coal supply volume of the i-th electricity market entity, is the planned thermal coal supply volume of the i-th electricity market entity.

[0244] In a further preferred solution, the division module clusters the observed data under multiple evaluation indicators to classify the electricity market entities, specifically including:

[0245] Using the Canopy clustering algorithm to conduct rough classification on the observed data under multiple evaluation indicators, and determining the number of clusters K and the initial cluster centers;

[0246] Taking the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of the electricity market entities using the K-means clustering algorithm.

[0247] In a further preferred solution, the division module uses the Canopy clustering algorithm to conduct rough classification on the observed data under multiple evaluation indicators, and determines the number of clusters K and the initial cluster centers. The steps include:

[0248] S111: Set initial distance thresholds T1 and T2 based on the observed data, where T1 > T2;

[0249] S112: Arbitrarily select a power market entity P, and calculate the shortest distance D between the power market entity P and all cluster centers respectively;

[0250] S113: If the shortest distance D is less than or equal to T2, then remove the power market entity P; if the shortest distance D is less than T1 and greater than T2, then add the power market entity P to the cluster corresponding to the cluster center with the shortest distance to it; if the shortest distance D is greater than or equal to T1, then the power market entity P forms a new cluster;

[0251] S114: Repeat all of the above S112 and S113 until all power market entities are classified.

[0252] In a further preferred solution, the partitioning module takes the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final type partitioning of power market entities using the K-means clustering algorithm. The steps include:

[0253] S121: Calculate the distances from each power market entity to the K cluster centers, and partition each power market entity into the class corresponding to the cluster center with the shortest distance;

[0254] S122: For each class, recalculate its cluster center;

[0255] S123: Repeat S121 and S122 until a preset termination condition is reached to obtain the final type partitioning of power market entities.

[0256] In a further preferred solution, the calculation formula for the cluster center in the partitioning module is:

[0257]

[0258] where a j is the new cluster center, |c i | is the number of power market entities in the c i th class, and t is the value of the power market entity belonging to this class.

[0259] In a further preferred solution, the objective weight acquisition module calculates the distances between different classes under the same evaluation index, and obtains the dispersion degree of each evaluation index based on the distances, specifically including:

[0260] Calculate the distances between any two cluster centers of each evaluation index;

[0261] The variance of all distances under the same evaluation index is used as the degree of dispersion of the evaluation index.

[0262] In a further optimized solution, the calculation formula for the degree of dispersion in the objective weight acquisition module is:

[0263]

[0264] where σ m is the degree of dispersion of the m-th evaluation index, and K m is the final number of clusters of the observed data of the m-th evaluation index, is the distance between the r-th cluster center and the j-th cluster center in the m-th evaluation index, is the average value of the distances of the cluster centers under the m-th evaluation index.

[0265] In a further optimized solution, the power system hierarchical structure model in the subjective weight acquisition module includes an objective layer, a criterion layer, and an index layer, where:

[0266] The objective layer is used to define the evaluation objective when allocating power resources;

[0267] The criterion layer is used to determine the main factors affecting the evaluation objective;

[0268] The index layer is used to determine the various evaluation indexes.

[0269] In a further optimized solution, the subjective weight acquisition module calculates the subjective weights of each evaluation index by performing calculations on the pre-constructed power system hierarchical structure model. The steps include:

[0270] Compare each element in the criterion layer pairwise to obtain the criterion layer judgment matrix; and compare each element in the index layer pairwise to obtain the index layer judgment matrix;

[0271] Perform a consistency test on the criterion layer judgment matrix and assign weights to obtain the weights of each element in the criterion layer, and perform a consistency test on the index layer judgment matrix and assign weights to obtain the weights of each element in the index layer;

[0272] Multiply the weights of each evaluation index relative to different indexes by the weights of the elements in the criterion layer and sum them to obtain the subjective weights of each evaluation index.

[0273] In a further optimized solution, when the subjective weight acquisition module performs a consistency test on the criterion layer judgment matrix and assigns weights to obtain the weights of each element in the criterion layer, and performs a consistency test on the index layer judgment matrix and assigns weights to obtain the weights of each element in the index layer, it includes:

[0274] Perform a consistency test on the judgment matrix of the criterion layer. If the test is passed, use the eigenvalue method, geometric mean method, or arithmetic mean method to calculate the weights of the elements in the criterion layer relative to the target layer.

[0275] Perform a consistency test on the judgment matrix of the index layer. If the test is passed, use the eigenvalue method, geometric mean method, or arithmetic mean method to calculate the weights of the elements in the index layer relative to the criterion layer.

[0276] In a further optimized solution, the calculation formula for the comprehensive score of the electricity market entity in the comprehensive score acquisition module is:

[0277]

[0278] where s i is the comprehensive score of the i-th electricity market entity, is the reporting accuracy rate of the operation parameters of the i-th electricity market entity, is the electricity quantity performance fulfillment rate of the i-th electricity market entity, is the auxiliary service fulfillment rate of the i-th electricity market entity, is the equipment operation reliability rate of the i-th electricity market entity, is the reliable rate of coal supply for electricity of the i-th electricity market entity, w1 is the weight coefficient of the reporting accuracy rate of operation parameters, w2 is the weight coefficient of the electricity quantity performance fulfillment rate, w3 is the weight coefficient of the auxiliary service fulfillment rate, w4 is the weight coefficient of the equipment operation reliability rate, and w5 is the weight coefficient of the reliable rate of coal supply for electricity.

[0279] In a further optimized solution, the calculation formula for the objective weight in the objective weight acquisition module is:

[0280]

[0281] where, is the objective weight of the m-th evaluation index, σ r is the degree of dispersion of the r-th evaluation index, N is the total number of evaluation indexes, and σ m is the degree of dispersion of the m-th evaluation index.

[0282] In a further optimized solution, the observed data in the partitioning module is obtained from various data sources using data mining and / or web crawler technology.

[0283] Embodiment 3

[0284] As Figure 8As shown, the present invention also provides an electronic device, which may be a computer device, a single-chip microcomputer device, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, the processor, and the transceiver component are connected through a bus; the memory can be used to store an execution program, and an exemplary execution program may include instructions; the processor is used to execute the instructions stored in the memory. The memory can also be used to store data, and this data can be called and / or modified when the instructions are executed.

[0285] The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of a data-driven power resource evaluation and configuration method in the above embodiment.

[0286] Embodiment 4

[0287] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device-readable storage medium (Memory). The electronic device-readable storage medium is a memory device in the electronic device, and is used to store programs and data. It can be understood that the storage medium here can include both the built-in storage medium in the electronic device, and of course can also include the extended storage medium supported by the electronic device. The storage medium provides a storage space, and this storage space stores the operating system of the terminal. And, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space. These instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory, or a non-volatile memory, such as at least one disk memory. By the processor loading and executing one or more instructions stored in the storage medium, the steps of a data-driven power resource evaluation and configuration method in the above embodiment can be implemented.

[0288] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0289] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or combinations of blocks.

[0290] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or combinations of blocks.

[0291] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or combinations of blocks.

[0292] The above are only embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A data-driven power resource evaluation and allocation method, characterized in that Including: Clustering the observed data under multiple evaluation indicators to classify the electricity market entities; Calculating the distances between different classes under the same evaluation indicator, obtaining the dispersion degree of each evaluation indicator based on the distances, and obtaining the objective weight of the corresponding evaluation indicator by taking the ratio of the dispersion degree of a single evaluation indicator to the sum of the dispersion degrees of all evaluation indicators; Calculating the subjective weight of each evaluation indicator by calculating the pre-constructed electricity system hierarchical structure model; Performing weighted summation on the objective weight and the subjective weight to obtain the comprehensive weight of each evaluation indicator; Performing weighted summation on each evaluation indicator with the comprehensive weight of each evaluation indicator as the coefficient to obtain the comprehensive score of the corresponding electricity market entity, and allocating electric power resources to the electricity market entity based on the comprehensive score.

2. The data-driven power resource evaluation and allocation method according to claim 1, wherein The evaluation indicators include: The accuracy rate of declared operation parameters, the completion rate of power quantity performance, the completion rate of auxiliary services, the reliable rate of equipment operation, or the reliable rate of coal supply.

3. The data-driven power resource evaluation and allocation method according to claim 2, wherein The calculation formula for the accuracy rate of declared operation parameters is: Among them, is the reporting accuracy rate of the operating parameters of the i-th electricity market entity, r i c is the number of incorrectly reported parameters of the i-th electricity market entity, r i y is the number of omitted reported parameters of the i-th electricity market entity, r i is the total number of reported parameters of the i-th electricity market entity.

4. The data-driven power resource evaluation and allocation method according to claim 2, characterized in that The calculation formula for the completion rate of power quantity performance is: Among them, is the electricity quantity compliance completion rate of the i-th electricity market entity, is the actual settled electricity quantity of the i-th electricity market entity, is the total winning bid electricity quantity of the i-th electricity market entity.

5. The data-driven power resource evaluation and allocation method according to claim 2, wherein The calculation formula for the completion rate of auxiliary services is: Among them, is the completion rate of the ancillary services of the i-th electricity market entity, is the actual power of the i-th electricity market entity, is the declared power of the i-th electricity market entity.

6. The data-driven power resource evaluation and allocation method according to claim 2, wherein The calculation formula for the reliable rate of equipment operation is: Among them, is the equipment operation reliability rate of the i-th electricity market entity, is the planned operation time of the i-th electricity market entity, is the unplanned outage time of the i-th electricity market entity.

7. The data-driven power resource evaluation and allocation method according to claim 2, wherein The calculation formula for the reliable rate of coal supply is: Among them, is the reliable rate of the thermal coal supply of the i-th electricity market entity, is the actual thermal coal supply volume of the i-th electricity market entity, is the planned thermal coal supply volume of the i-th electricity market entity.

8. The data-driven power resource evaluation and allocation method according to claim 1, wherein The clustering of the observed data under multiple evaluation indicators to classify the electricity market entities specifically includes: Coarsely classifying the observed data under multiple evaluation indicators by using the Canopy clustering algorithm to determine the number of clusters K and the initial cluster centers; Taking the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of the electricity market entities by using the K-means clustering algorithm.

9. The data-driven power resource evaluation and allocation method according to claim 8, wherein, The steps of coarsely classifying the observed data under multiple evaluation indicators by using the Canopy clustering algorithm to determine the number of clusters K and the initial cluster centers include: S111: Setting initial distance thresholds T1 and T2 based on the observed data, where T1>T2; S112: Arbitrarily selecting an electricity market entity P and calculating the nearest distance D between the electricity market entity P and all cluster centers respectively; S113: If the nearest distance D is less than or equal to T2, then the electricity market entity P is removed; if the nearest distance D is less than T1 and greater than T2, then the electricity market entity P is added to the cluster corresponding to the nearest cluster center; if the nearest distance D is greater than or equal to T1, then the electricity market entity P forms a new cluster; S114: Repeating all of the above S112 and S113 until all the electricity market entities are classified.

10. The data-driven power resource evaluation and allocation method according to claim 9, wherein The steps of taking the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of the electricity market entities by using the K-means clustering algorithm include: S121: Calculating the distances between each electricity market entity and the K cluster centers, and dividing each electricity market entity into the class corresponding to the cluster center with the smallest distance; S122: Recalculating the cluster center for each category; S123: Repeating step 121 and step 122 until a preset termination condition is reached to obtain the final classification of the electricity market entities.

11. The data-driven power resource evaluation and allocation method according to claim 10, wherein The calculation formula for the cluster center in S122 is as follows: Among them, a j is the new clustering center, |c i | is the number of power market entities in the c i -th class, and t is the value of the power market entities belonging to this class.

12. The data-driven power resource evaluation and allocation method according to claim 1, wherein Calculating the distances between different classes under the same evaluation index, and obtaining the dispersion degree of each evaluation index based on the distances, specifically including: Calculating the distances between any two cluster centers of each evaluation index; Using the variance of all distances under the same evaluation index as the dispersion degree of the evaluation index.

13. The data-driven power resource evaluation and allocation method according to claim 1 or 12, characterized in that The calculation formula for the dispersion degree is as follows: Among them, σ m is the degree of dispersion of the m-th evaluation index, and K m is the final number of clusters of the observed data of the m-th evaluation index, is the distance between the r-th cluster center and the j-th cluster center in the m-th evaluation index, is the average value of the distances between the cluster centers under the m-th evaluation index.

14. The data-driven power resource evaluation and allocation method according to claim 1, characterized in that The power system hierarchical structure model includes an objective layer, a criterion layer, and an index layer, where: The objective layer is used to define the evaluation objective when allocating power resources; The criterion layer is used to determine the main factors affecting the evaluation objective; The index layer is used to determine the multiple evaluation indexes.

15. The data-driven power resource evaluation and allocation method according to claim 14, wherein Calculating the subjective weights of each evaluation index by calculating the pre-constructed power system hierarchical structure model, the steps include: Making pairwise comparisons of the elements in the criterion layer to obtain the criterion layer judgment matrix; and making pairwise comparisons of the elements in the index layer to obtain the index layer judgment matrix; Conducting a consistency test on the criterion layer judgment matrix and assigning weights to obtain the weights of the elements in the criterion layer, and conducting a consistency test on the index layer judgment matrix and assigning weights to obtain the weights of the elements in the index layer; Multiplying the weights of each evaluation index relative to different indexes by the weights of the elements in the criterion layer and summing them to obtain the subjective weights of each evaluation index.

16. The data-driven power resource evaluation and allocation method according to claim 15, wherein Conducting a consistency test on the criterion layer judgment matrix and assigning weights to obtain the weights of the elements in the criterion layer, and conducting a consistency test on the index layer judgment matrix and assigning weights to obtain the weights of the elements in the index layer, including: Conducting a consistency test on the criterion layer judgment matrix. If the test is passed, the eigenvalue method, geometric mean method, or arithmetic mean method is used to calculate the weights of the elements in the criterion layer relative to the objective layer; Conducting a consistency test on the index layer judgment matrix. If the test is passed, the eigenvalue method, geometric mean method, or arithmetic mean method is used to calculate the weights of the elements in the index layer relative to the criterion layer.

17. The data-driven power resource evaluation and allocation method according to claim 2, wherein The calculation formula for the comprehensive score of the power market entity is as follows: where s i is the comprehensive score of the i-th electricity market entity, is the reporting accuracy rate of the operation parameters of the i-th electricity market entity, is the electricity quantity compliance completion rate of the i-th electricity market entity, is the auxiliary service completion rate of the i-th electricity market entity, is the equipment operation reliability rate of the i-th electricity market entity, is the reliable rate of thermal coal supply of the i-th electricity market entity. w1 is the weight coefficient of the reporting accuracy rate of operation parameters, w2 is the weight coefficient of the electricity quantity compliance completion rate, w3 is the weight coefficient of the auxiliary service completion rate, w4 is the weight coefficient of the equipment operation reliability rate, and w5 is the weight coefficient of the reliable rate of thermal coal supply.

18. The data-driven power resource evaluation and allocation method according to claim 1, wherein The calculation formula for the objective weight is as follows: wherein, is the objective weight of the m-th evaluation index, and σ r is the degree of dispersion of the r-th evaluation index, N is the total number of evaluation indexes, and σ m is the degree of dispersion of the m-th evaluation index.

19. The data-driven power resource evaluation and allocation method according to claim 1, wherein The observed data is obtained from various data sources by using data mining and / or web crawling technology.

20. A data-driven power resource evaluation and allocation system, characterized in that, Including: A partitioning module: used to cluster the observed data under multiple evaluation indexes to classify the power market entities; An objective weight obtaining module: used to calculate the distances between different classes under the same evaluation index, obtain the dispersion degree of each evaluation index based on the distances, and obtain the objective weight of the corresponding evaluation index by taking the ratio of the dispersion degree of a single evaluation index to the sum of the dispersion degrees of all evaluation indexes; A subjective weight obtaining module: used to calculate the subjective weights of each evaluation index by calculating the pre-constructed power system hierarchical structure model; A comprehensive weight obtaining module: used to perform weighted summation on the objective weight and the subjective weight to obtain the comprehensive weight of each evaluation index; A comprehensive score obtaining module: used to perform weighted summation on each evaluation index with the comprehensive weight of each evaluation index as the coefficient to obtain the comprehensive score of the corresponding power market entity, and allocate power resources to the power market entity based on the comprehensive score.

21. The data-driven power resource evaluation and allocation system according to claim 20, wherein The evaluation indexes in the partitioning module include: The declaration accuracy rate of operation parameters, the fulfillment rate of power quantity, the fulfillment rate of ancillary services, the reliable operation rate of equipment, or the reliable supply rate of thermal coal.

22. The data-driven power resource evaluation and allocation system according to claim 21, wherein The calculation formula for the declaration accuracy rate of operation parameters in the division module is: Among them, is the reporting accuracy rate of the operation parameters of the i-th electricity market entity, r i c is the number of incorrectly reported parameters of the i-th electricity market entity, r i y is the number of omitted reported parameters of the i-th electricity market entity, r i is the total number of reported parameters of the i-th electricity market entity.

23. The data-driven power resource evaluation and allocation system according to claim 21, wherein The calculation formula for the fulfillment rate of power quantity in the division module is: Among them, is the electricity quantity compliance completion rate of the i-th electricity market entity, is the actual settlement electricity quantity of the i-th electricity market entity, is the total winning bid electricity quantity of the i-th electricity market entity.

24. The data-driven power resource evaluation and allocation system according to claim 21, wherein The calculation formula for the fulfillment rate of ancillary services in the division module is: Among them, is the auxiliary service completion rate of the i-th electricity market entity, is the actual power of the i-th electricity market entity, is the declared power of the i-th electricity market entity.

25. The data-driven power resource evaluation and allocation system according to claim 21, wherein The calculation formula for the reliable operation rate of equipment in the division module is: wherein, is the equipment operation reliability rate of the i-th electricity market entity, is the planned operation time of the i-th electricity market entity, is the unplanned outage time of the i-th electricity market entity.

26. The data-driven power resource evaluation and allocation system according to claim 21, wherein The calculation formula for the reliable supply rate of thermal coal in the division module is: Among them, is the reliable rate of thermal coal supply for the i-th electricity market entity, is the actual thermal coal supply volume of the i-th electricity market entity, is the planned thermal coal supply volume of the i-th electricity market entity.

27. The data-driven power resource evaluation and allocation system according to claim 20, wherein The division module clusters the observed data under multiple evaluation indicators to classify the electricity market entities, specifically including: Using the Canopy clustering algorithm to roughly classify the observed data under multiple evaluation indicators, and determining the number of clusters K and the initial cluster centers; Taking the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of electricity market entities using the K-means clustering algorithm.

28. The data-driven power resource evaluation and allocation system according to claim 27, wherein The division module uses the Canopy clustering algorithm to roughly classify the observed data under multiple evaluation indicators, and determines the number of clusters K and the initial cluster centers. The steps include: S111: Based on the observed data, set the initial distance thresholds T1 and T2, where T1 > T2; S112: Arbitrarily select an electricity market entity P, and calculate the nearest distance D between the electricity market entity P and all cluster centers respectively; S113: If the nearest distance D is less than or equal to T2, then the electricity market entity P is excluded; if the nearest distance D is less than T1 and greater than T2, then the electricity market entity P is added to the cluster where the nearest cluster center is located; if the nearest distance D is greater than or equal to T1, then the electricity market entity P forms a new cluster; S114: Repeat all of the above S112 and S113 until all electricity market entities are classified.

29. The data-driven power resource evaluation and allocation system according to claim 28, wherein The division module takes the number of clusters K and the initial cluster centers as the input of the K-means clustering algorithm to perform the final classification of electricity market entities using the K-means clustering algorithm. The steps include: S121: Calculate the distances from each electricity market entity to the K cluster centers, and divide each electricity market entity into the class corresponding to the cluster center with the smallest distance; S122: For each class, recalculate its cluster center; S123: Repeat S121 and S122 until a preset termination condition is reached to obtain the final classification of electricity market entities.

30. The data-driven power resource evaluation and allocation system according to claim 29, wherein The calculation formula for the cluster center in the division module is: Among them, a j is the new clustering center, |c i | is the number of power market entities in the c i -th class, and t is the value of the power market entities belonging to this class.

31. The data-driven power resource evaluation and allocation system according to claim 20, wherein The objective weight acquisition module calculates the distances between different classes under the same evaluation indicator, and obtains the dispersion degree of each evaluation indicator based on the distances, specifically including: Calculating the distances between any two cluster centers of each evaluation indicator; Using the variance of all distances under the same evaluation indicator as the dispersion degree of the evaluation indicator.

32. The data-driven power resource evaluation and allocation system according to claim 20 or 31, characterized in that The calculation formula for the dispersion degree in the objective weight acquisition module is: Among them, σ m is the degree of dispersion of the m-th evaluation index, K m is the final number of clusters of the observed data of the m-th evaluation index, is the distance between the r-th cluster center and the j-th cluster center in the m-th evaluation index, is the average value of the distances between the cluster centers under the m-th evaluation index.

33. The data-driven power resource evaluation and allocation system according to claim 20, wherein, The power system hierarchical structure model in the subjective weight acquisition module includes an objective layer, a criterion layer, and an index layer, where: The objective layer is used to define the evaluation objective when allocating electric power resources; The criterion layer is used to determine the main factors affecting the evaluation objective; The index layer is used to determine the multiple evaluation indexes.

34. The data-driven power resource evaluation and allocation system according to claim 33, wherein The subjective weight acquisition module calculates the subjective weights of each evaluation index by performing calculations on a pre-constructed hierarchical structure model of the power system. The steps include: Making pairwise comparisons of the elements in the criterion layer to obtain the criterion layer judgment matrix; and making pairwise comparisons of the elements in the index layer to obtain the index layer judgment matrix; Performing consistency tests on the criterion layer judgment matrix and assigning weights to obtain the weights of the elements in the criterion layer, and performing consistency tests on the index layer judgment matrix and assigning weights to obtain the weights of the elements in the index layer; Multiplying the weights of each evaluation index relative to different indexes by the weights of the elements in the criterion layer and summing them to obtain the subjective weights of each evaluation index.

35. The data-driven power resource evaluation and allocation system according to claim 34, wherein The subjective weight acquisition module performs consistency tests on the criterion layer judgment matrix and assigns weights to obtain the weights of the elements in the criterion layer, and performs consistency tests on the index layer judgment matrix and assigns weights to obtain the weights of the elements in the index layer, including: Performing a consistency test on the criterion layer judgment matrix. If the test is passed, the eigenvalue method, geometric mean method or arithmetic mean method is used to calculate the weights of the elements in the criterion layer relative to the target layer; Performing a consistency test on the index layer judgment matrix. If the test is passed, the eigenvalue method, geometric mean method or arithmetic mean method is used to calculate the weights of the elements in the index layer relative to the criterion layer.

36. The data-driven power resource evaluation and allocation system according to claim 21, wherein The calculation formula for the comprehensive score of the power market entity in the comprehensive score acquisition module is: where s i is the comprehensive score of the i-th electricity market entity, is the accuracy rate of the declared operation parameters of the i-th electricity market entity, is the electricity quantity compliance completion rate of the i-th electricity market entity, is the ancillary service completion rate of the i-th electricity market entity, is the equipment operation reliability rate of the i-th electricity market entity, is the reliable rate of coal supply for the i-th electricity market entity, w1 is the weight coefficient of the accuracy rate of declared operation parameters, w2 is the weight coefficient of the electricity quantity compliance completion rate, w3 is the weight coefficient of the ancillary service completion rate, w4 is the weight coefficient of the equipment operation reliability rate, and w5 is the weight coefficient of the reliable rate of coal supply.

37. The data-driven power resource evaluation and allocation system according to claim 20, wherein The calculation formula for the objective weight in the objective weight acquisition module is: Among them, is the objective weight of the m-th evaluation index, and σ r is the degree of dispersion of the r-th evaluation index, N is the total number of evaluation indexes, and σ m is the degree of dispersion of the m-th evaluation index.

38. The data-driven power resource evaluation and allocation system according to claim 20, wherein In the partitioning module, the observed data is obtained from various data sources by using data mining and / or web crawler technology.

39. A computer device, characterized in that, Including: At least one processor and a memory; The memory and the processor are connected by a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the data-driven power resource evaluation and configuration method according to any one of claims 1 to 19 is implemented.

40. A computer-readable storage medium, characterized in that, There is an execution program stored thereon. When the execution program is executed, the data-driven power resource evaluation and configuration method according to any one of claims 1 to 19 is implemented.

Images