A new energy station voltage support performance evaluation method based on dynamic response and fuzzy C-means clustering

Abstract

This invention discloses a method for evaluating the voltage support performance of renewable energy power plants based on dynamic response and fuzzy C-means clustering. The method includes: acquiring several evaluation index data for renewable energy power plants, including static potential indicators and dynamic performance indicators; combining and weighting the evaluation index data to obtain a comprehensive weight vector for each indicator; inputting the comprehensive weight vector into a fuzzy C-means clustering (FCM) model for iterative calculation to obtain a membership matrix indicating the different performance levels of each power plant; and outputting the voltage support performance evaluation result based on the membership matrix. This invention provides more accurate quantitative basis for the planning and design, grid connection review, operation status monitoring, and control strategy optimization of renewable energy power plants, helping to improve the stability of high-proportion renewable energy grid connections.

Description

Technical Field

[0001] This invention relates to the field of power system analysis and evaluation technology, and in particular to a method for evaluating the voltage support performance of new energy power stations based on dynamic response and fuzzy C-means clustering. Background Technology

[0002] With the advancement of the "dual carbon" target, new energy sources, represented by wind power and photovoltaics, are being connected to the grid on a large scale. However, most new energy power generation units are connected to the grid through power electronic converters, whose dynamic response characteristics differ fundamentally from those of traditional synchronous generators. When a grid fault causes a voltage drop, the ability of new energy power plants to provide timely and effective reactive power support is crucial for preventing large-scale disconnection of new energy from the grid and ensuring grid voltage stability.

[0003] Currently, the assessment of the voltage support capability of new energy sources is mostly based on static system parameters. However, these methods have the following significant limitations:

[0004] 1. Static Assessment: Most existing methods rely solely on static parameters, leading to a significant disconnect between assessment results and actual fault response. For example, a power station may have an SVG capacity of 50 Mvar, but its control system response delay reaches 300 ms, failing to provide support in the initial stage of a fault (<100 ms), and its actual support capability is far lower than the assessed value. Therefore, static assessment can no longer meet the refined safety control requirements of power grids with a high proportion of renewable energy.

[0005] 2. Model limitations: Some evaluation methods use hard classification models such as grey clustering, which forcibly divide the stations into discrete levels such as "strong", "relatively strong", and "average". This makes it difficult to scientifically handle the ambiguity of continuous changes in the support capacity between different stations, and the evaluation results may not be accurate enough.

[0006] 3. Lack of innovation in technology combinations: Although existing technologies have proposed dynamic response indices or fuzzy clustering methods, there are no reports on a technical solution that combines dynamic response indices with fuzzy C-means clustering algorithms for evaluating the voltage support performance of new energy sources. Such a combination can solve the fuzziness and dynamism problems in voltage support performance evaluation.

[0007] Therefore, existing technologies are insufficient to accurately and comprehensively assess the actual voltage support performance of renewable energy power plants, and cannot meet the needs of refined analysis of high-proportion renewable energy power grids. Summary of the Invention

[0008] To address the technical problems existing in the prior art, this invention proposes a method for evaluating the voltage support performance of new energy power stations based on dynamic response and fuzzy C-means clustering. This method aims to shift the evaluation from "static capability" to "dynamic performance." By introducing dynamic indicators that reflect the actual response speed and employing advanced clustering algorithms that can handle fuzziness, it achieves a more accurate and objective quantitative evaluation of the voltage support performance of new energy power stations, providing a reliable basis for the planning, grid connection evaluation, and operation optimization of new energy power stations.

[0009] To achieve the above objectives, this invention provides a method for evaluating the voltage support performance of new energy power stations based on dynamic response and fuzzy C-means clustering, comprising:

[0010] Acquire data on several evaluation indicators for new energy power plants, including static potential indicators and dynamic performance indicators.

[0011] The evaluation index data are combined and weighted to obtain the comprehensive weight vector of each index;

[0012] The comprehensive weight vector is input into the fuzzy C-means clustering (FCM) model for iterative calculation to obtain the membership matrix of each station belonging to different performance levels.

[0013] The output voltage support performance evaluation results are based on the membership matrix.

[0014] Preferably, the static potential indicators include the capacity of the new energy unit, the capacity of the static var generator (SVG), the reactive current ratio coefficient, the short-circuit ratio, and the impedance ratio; the dynamic performance indicators include the reactive current response time and the available energy storage capacity, wherein the reactive current response time is the time required from the detection of a voltage drop at the grid connection point to the actual reactive current output by the new energy unit and the SVG reaching the preset command value ratio; and the available energy storage capacity is the effective reactive power capacity that the energy storage system in the power station can provide during a fault.

[0015] Preferably, the evaluation index data are combined and weighted to obtain the comprehensive weight of each index, including:

[0016] The subjective weight vector is calculated using the analytic hierarchy process (AHP).

[0017] The objective weight vector is calculated using the entropy weight method;

[0018] With the goal of minimizing the sum of squared deviations between the combined weights and the subjective and objective weights, an optimization model is established and solved to obtain the comprehensive weights.

[0019] Preferably, the calculation of the subjective weight vector using the analytic hierarchy process includes:

[0020] Based on pairwise importance comparisons of various indicators by several power system experts, a judgment matrix is ​​constructed. The eigenvector corresponding to the largest eigenvalue of the judgment matrix is ​​calculated and normalized to obtain the subjective weight vector u = [u1, u2, ..., u7].

[0021] The objective weight vector is calculated using the entropy weight method, including:

[0022] The original evaluation index data were dimensionless, and the entropy value e of each index was calculated. j and information utility value d j This allows us to obtain the objective weight vector v = [v1, v2, ..., v7];

[0023] The optimization model is as follows:

[0024]

[0025] The constraints for optimizing the model are:

[0026]

[0027] In the formula, η j The combined weights to be determined are: u j The subjective weights of each indicator;

[0028] Preferably, the comprehensive weights of each indicator are input into the fuzzy C-means clustering (FCM) model for iterative calculation to obtain the membership matrix of each station belonging to different performance levels, including:

[0029] Initialize the number of clusters, fuzzy index, and convergence threshold; randomly initialize the membership matrix.

[0030] Calculate the cluster center for each indicator data based on the current membership matrix and comprehensive weight vector;

[0031] Iteratively update the cluster centers and membership matrix, calculate the change in the membership matrix, and stop iterating when the change in membership is less than a preset value.

[0032] Preferably, the initialization of the cluster number, fuzzy index, and convergence threshold, and the random initialization of the membership matrix, include:

[0033] Set the number of clusters c=5, corresponding to performance levels of "weak", "weak", "average", "strong", and "strong"; set the fuzzy index m=2; set the convergence threshold ε; and randomly initialize the membership matrix. Where n is the number of new energy power stations to be evaluated. Let the initial membership degree of the i-th station belong to the k-th cluster be such that it satisfies and

[0034] Based on the current membership matrix U (t) Combined with the weight vector η, the cluster center of each indicator data is calculated as follows:

[0035]

[0036] In the formula, x is the cluster center. i =[x i1 ,x i2 ,...,x i7 ] represents the 7-dimensional index vector of the i-th station, and ⊙ represents element-wise multiplication. Let m be the membership degree of the i-th power station belonging to the k-th cluster, m be the fuzzy index, and n be the number of new energy power stations to be evaluated.

[0037] Preferably, the membership matrix is ​​updated using the following method:

[0038]

[0039] In the formula, Let c be the updated membership degree of the i-th station belonging to the k-th cluster, and c be the number of clusters. These are the k-th and j-th new cluster centers, respectively.

[0040] Preferably, the change in the membership matrix is ​​calculated as follows:

[0041]

[0042] In the formula, Δ (t+1) This represents the change in the membership matrix.

[0043] Preferably, the fuzzy index m of the fuzzy C-means clustering FCM model is adaptively adjusted based on the standard deviation of the dynamic performance class index, specifically as follows:

[0044]

[0045] Where, σ d σ represents the standard deviation of dynamic performance indicators. s The standard deviation is the static potential index.

[0046] Preferably, the voltage support performance evaluation results include: the final voltage support performance level, fuzzy information, and a fuzzy early warning mechanism;

[0047] The final voltage support performance level is determined by the maximum membership degree, specifically:

[0048]

[0049] In the formula, L i The voltage support level is the final level, and k is the corresponding performance level. k=1 corresponds to "weak", k=2 corresponds to "relatively weak", k=3 corresponds to "average", k=4 corresponds to "relatively strong", and k=5 corresponds to "strong".

[0050] When the maximum membership value is lower than the preset threshold, a performance fuzzy warning is triggered.

[0051] Compared with the prior art, the present invention has the following advantages and technical effects:

[0052] (1) More accurate assessment: This invention assesses the "dynamic support performance" of new energy power plants rather than their static potential by introducing dynamic indicators such as "reactive current response time" and "energy storage callable capacity".

[0053] (2) More scientific processing: The present invention adopts the fuzzy C-means clustering (FCM) algorithm, which utilizes its membership output characteristics to scientifically quantify the fuzziness and uncertainty in the evaluation process, avoids the information loss caused by hard classification, and makes the evaluation results more objective and reasonable.

[0054] (3) More comprehensive results: The evaluation results of the present invention not only include the final performance level, but also provide the maximum membership degree as a confidence index, making the evaluation results more informative and valuable for decision-making. In particular, the fuzzy warning mechanism triggered when the maximum membership degree is lower than the preset threshold provides decision-makers with additional risk warnings.

[0055] (4) Outstanding innovation: This invention does not simply use FCM for classification, but rather innovatively combines dynamic indicators with the FCM algorithm to solve the specific technical problem of fuzziness in dynamic performance evaluation, resulting in a synergistic technical effect with quantifiable confidence.

[0056] (5) More practical: The method of this invention provides more accurate quantitative basis for the planning and design of new energy power plants, grid connection review, operation status monitoring and control strategy optimization, which helps to improve the stability of high proportion of new energy access to the grid. Attached Figure Description

[0057] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0058] Figure 1 This is a flowchart of a method for evaluating the voltage support performance of new energy power stations based on dynamic response and fuzzy C-means clustering, according to an embodiment of the present invention.

[0059] Figure 2This is a schematic diagram of the "static-dynamic" comprehensive evaluation index system constructed according to an embodiment of the present invention. Detailed Implementation

[0060] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0061] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0062] This embodiment proposes a method for evaluating the voltage support performance of new energy power stations based on dynamic response and fuzzy C-means clustering, including:

[0063] Acquire data on several evaluation indicators for new energy power plants, including static potential indicators and dynamic performance indicators.

[0064] The evaluation index data are combined and weighted to obtain the comprehensive weight vector of each index;

[0065] The comprehensive weight vector is input into the fuzzy C-means clustering (FCM) model for iterative calculation to obtain the membership matrix of each station belonging to different performance levels.

[0066] The output voltage support performance evaluation results are based on the membership matrix.

[0067] Specifically, to address the issue that existing evaluation methods, which are mostly based on static parameters, fail to reflect the actual dynamic response performance of renewable energy power plants during faults, this embodiment constructs a comprehensive evaluation index system that includes both static potential and dynamic performance. Two new dynamic indicators, reactive current response time and available energy storage capacity, are added. A combined weighting method is used to determine the weights of each indicator. The traditional hard classification model is abandoned, and a fuzzy C-means clustering (FCM) algorithm is introduced. Through iterative calculation, the membership degree of each power plant to different performance levels is obtained. Finally, the voltage support performance level of each renewable energy power plant is determined based on the maximum membership degree, and the membership degree value is output as the confidence level of the evaluation result. This emphasizes the synergistic application of dynamic indicators and the FCM algorithm, solving the fuzziness and dynamism problems in voltage support performance evaluation. It produces a synergistic technical effect with quantifiable confidence levels, enabling more accurate and objective quantitative evaluation of the dynamic voltage support performance of renewable energy power plants, and providing a decision-making basis for the safe and stable operation of the power grid.

[0068] Furthermore, static potential indicators include: new energy unit capacity (x1), static var generator (SVG) capacity (x2), reactive current proportional gain (x3), short-circuit ratio (x4), and impedance ratio (x5);

[0069] Dynamic performance metrics include: reactive current response time (x6) and available energy storage capacity (x7);

[0070] Among them, reactive current response time (x6) refers to the time required from the detection of voltage drop at the grid connection point to the actual reactive current output by the new energy unit and SVG reaching 90% of its command value, in milliseconds (ms); energy storage callable capacity (x7) refers to the effective reactive power capacity that the energy storage system in the site can provide during a fault, in megavars (Mvar).

[0071] Specifically, the introduction of reactive current response time (x6) and available energy storage capacity (x7) specifically addresses the technical problem that static assessment in existing technologies cannot reflect dynamic response performance during faults. Furthermore, dynamic performance indicators and static potential indicators are combined and weighted to form an organic whole, working together on the input of the FCM algorithm to produce a synergistic technical effect of dynamic response characteristics and supporting potential collaborative assessment.

[0072] Furthermore, the overall weight of each indicator is obtained, including:

[0073] The subjective weight vector is calculated using the analytic hierarchy process (AHP).

[0074] The objective weight vector is calculated using the entropy weight method;

[0075] With the goal of minimizing the sum of squared deviations between the combined weights and the subjective and objective weights, an optimization model is established and solved to obtain the comprehensive weights.

[0076] Specifically, including:

[0077] Subjective weight calculation: The Analytic Hierarchy Process (AHP) is adopted. At least three experts in the field of power system are invited to conduct pairwise importance comparisons of each indicator to construct a judgment matrix. The eigenvector corresponding to the largest eigenvalue of the judgment matrix is ​​calculated and normalized to obtain the subjective weight vector u = [u1, u2, ..., u7].

[0078] Objective weight calculation: Using the entropy weight method, the original evaluation index data is first processed to be dimensionless, and the entropy value e of each index is calculated. j and information utility value d j This leads to the objective weight vector v = [v1, v2, ..., v7];

[0079] Combined weight calculation: With the objective of minimizing the sum of squared deviations between the combined weights and the subjective and objective weights, the following optimization model is established and solved:

[0080]

[0081] The constraints are:

[0082]

[0083] In the formula, η j The combined weights to be determined are: u j The subjective weights of each indicator;

[0084] The optimization model is solved using the least squares method to obtain the final combined weight vector η = [η1, η2, ..., η7]. The combined weight calculation is specifically optimized for the dynamic response fuzziness problem in voltage support performance evaluation to ensure that dynamic performance indicators receive appropriate weights in the evaluation.

[0085] In the process of calculating the combined weight vector η, weight constraints are set for dynamic performance indicators to ensure that η6+η7≥0.25, thereby ensuring that the dynamic response indicators receive sufficient weight in the evaluation and solving the technical problem that the importance of dynamic indicators is underestimated in the existing technology.

[0086] Furthermore, the comprehensive weights of each indicator are input into the fuzzy C-means clustering (FCM) model for iterative calculation to obtain the membership matrix of each station belonging to different performance levels, including:

[0087] Initialize the number of clusters, fuzzy index, and convergence threshold; randomly initialize the membership matrix.

[0088] Calculate the cluster center for each indicator data based on the current membership matrix and comprehensive weight vector;

[0089] Iteratively update the cluster centers and membership matrix, calculate the change in the membership matrix, and stop iterating when the change in membership is less than a preset value.

[0090] Specifically, including:

[0091] Model initialization: Set the number of clusters c=5, corresponding to the performance levels "weak", "relatively weak", "average", "relatively strong" and "strong";

[0092] Set the fuzziness index m = 2; set the convergence threshold ε;

[0093] Randomly initialize the membership matrix:

[0094]

[0095] Where n is the number of new energy power stations to be evaluated. This represents the initial membership degree of the i-th station to the k-th cluster, satisfying... and

[0096] Iterative calculations include:

[0097] (a) Update cluster centers: based on the current membership matrix U (t) Combine the weight vector η to calculate the cluster center.

[0098]

[0099] In the formula, x i =[x i1 ,x i2 ,...,x i7 ] represents the 7-dimensional index vector of the i-th station, and ⊙ represents element-wise multiplication. Let η be the membership degree of the i-th power station belonging to the k-th cluster, η be the combined weight vector, m be the fuzzy index, and n be the number of new energy power stations to be evaluated.

[0100] (b) Update the membership matrix: based on the new cluster centers. Calculate the membership degree of each station i to each cluster k.

[0101]

[0102] In the formula, Let c be the updated membership degree of the i-th station belonging to the k-th cluster, and c be the number of clusters. These are the k-th and j-th new cluster centers, respectively.

[0103] (c) Convergence criterion: Calculate the change in the membership matrix:

[0104]

[0105] If Δ (t+1) If ε < 1, the algorithm converges and the iteration stops; otherwise, let t = t + 1 and return to step (a) to continue the iteration.

[0106] In this embodiment, the FCM algorithm is specifically optimized for the ambiguity of voltage support performance evaluation. By synergistically weighting dynamic and static indicators, it solves the technical problem that the traditional FCM algorithm cannot accurately reflect dynamic response characteristics in power system evaluation.

[0107] Furthermore, the fuzzy exponent m of the fuzzy C-means clustering FCM model is adaptively adjusted based on the standard deviation of the dynamic performance class index, specifically as follows:

[0108]

[0109] Where, σ d σ represents the standard deviation of dynamic performance indicators. s The standard deviation is the static potential index.

[0110] Specifically, the application of the fuzzy C-means clustering (FCM) algorithm is specifically optimized for the fuzzy problem in the performance evaluation of new energy voltage support. By incorporating the dynamic response index into the FCM input vector, it produces an effective technical effect of synergistic effect between dynamic index and fuzzy classification.

[0111] Furthermore, the voltage support performance evaluation results include: the final voltage support performance level, fuzzy information, and a fuzzy early warning mechanism;

[0112] The final voltage support performance level is determined by the maximum membership degree, specifically:

[0113]

[0114] In the formula, L i The voltage support level is the final level, and k is the corresponding performance level. k=1 corresponds to "weak", k=2 corresponds to "relatively weak", k=3 corresponds to "average", k=4 corresponds to "relatively strong", and k=5 corresponds to "strong".

[0115] When the maximum membership value is lower than the preset threshold, a performance fuzzy warning is triggered.

[0116] Specifically, the output voltage support performance evaluation results include:

[0117] S4.1 Determine the final level: For each new energy power station i, its voltage support performance level is determined by the maximum membership degree;

[0118] S4.2, Provide fuzzy information: Simultaneously output the maximum membership value of this station. As a confidence level of the evaluation result; the maximum membership value is specifically used to quantify the degree of certainty of the evaluation result, which solves the technical problem that traditional voltage support performance evaluation cannot provide the confidence level of the evaluation result;

[0119] S4.3 Fuzzy warning mechanism: When the maximum membership value is lower than the preset threshold τ, a performance fuzzy warning is triggered, where the value of τ is in the range of 0.5≤τ≤0.7;

[0120] The fuzzy early warning mechanism provides technical early warnings specifically for situations where voltage support performance is at fuzzy boundaries, offering additional risk alerts for power grid dispatching decisions and solving the technical problem of existing technologies being unable to identify fuzzy boundary stations.

[0121] To more clearly illustrate the technical solution of the present invention, specific embodiments are provided below for description:

[0122] like Figure 1 As shown, the implementation process of the method described in this embodiment includes four main steps.

[0123] Step S1: Obtain data on multiple evaluation indicators for new energy power plants;

[0124] like Figure 2 The indicator system in this embodiment is divided into two layers. The criterion layer includes "static potential" and "dynamic performance".

[0125] The indicator layer contains seven specific indicators:

[0126] Static potential: new energy unit capacity x1, SVG capacity x2, reactive current proportional coefficient x3, short-circuit ratio x4, impedance ratio x5;

[0127] Dynamic performance: reactive current response time x6, available energy storage capacity x7;

[0128] This embodiment collected data on the above seven indicators from 10 renewable energy power stations in a certain regional power grid, as shown in Table 1:

[0129] Table 1

[0130] Station Number <![CDATA[x1(MW)]]> <![CDATA[x2(Mvar)]]> <![CDATA[x3]]> <![CDATA[x4]]> <![CDATA[x5]]> <![CDATA[x6(ms)]]> <![CDATA[x7(Mvar)]]> A01 100 30 0.95 3.5 4.2 80 15 A02 80 20 0.85 2.8 3.6 250 5 A03 120 40 0.92 4.2 5.0 120 20 A04 60 15 0.78 2.2 2.8 320 3 A05 90 25 0.88 3.2 3.9 150 10 A06 110 35 0.90 3.8 4.5 100 18 A07 70 18 0.82 2.5 3.2 280 6 A08 95 28 0.89 3.4 4.1 180 12 A09 85 22 0.86 3.0 3.7 220 8 A10 105 32 0.93 4.0 4.7 90 16

[0131] Step S2: Determine the combined weights of each evaluation indicator:

[0132] 1. Subjective Weighting: Five power system experts were invited to conduct pairwise comparisons of the importance of each indicator to voltage support performance, constructing a judgment matrix. By calculating and normalizing the eigenvectors, the average subjective weight vector u = [0.14, 0.17, 0.11, 0.21, 0.09, 0.16, 0.12] was obtained.

[0133] 2. Objective weights: The original data is normalized (using min-max standardization) to eliminate the influence of dimensions.

[0134] Then, according to the entropy weight method calculation formula, the objective weight vector is obtained:

[0135] v=[0.16,0.19,0.10,0.18,0.11,0.14,0.12],

[0136] 3. Combined weights: Substitute the average subjective weight vector u and the objective weight vector v into the optimization model, and solve using the least squares method to finally obtain the combined weight vector η = [0.15, 0.18, 0.11, 0.20, 0.10, 0.15, 0.11].

[0137] Step S3: Establish a performance evaluation model based on fuzzy C-means clustering;

[0138] 1. Multiply the 7-dimensional index data of the 10 stations by the combined weight vector η and add them.

[0139] 2. Set c = 5, m = 2, ε = 1e-6;

[0140] 3. Randomly initialize the membership matrix U;

[0141] 4. Iteratively execute the two steps of "updating cluster centers" and "updating membership matrix" until the algorithm converges, and obtain the final membership matrix U and cluster centers.

[0142] Step S4: Output voltage support performance evaluation results;

[0143] Each site was evaluated based on the final membership matrix U. The evaluation results for the 10 sites are shown in Table 2.

[0144] Table 2

[0145]

[0146] As can be seen from the evaluation results in Table 2, most sites received clear performance ratings with very high maximum membership values ​​(confidence levels). For example, site A04 has a membership of 0.99 in the "Strong" category, and site A08 has a membership of 0.99 in the "Moderate" category. This indicates that the method in this embodiment can provide highly confident evaluations of sites with significant performance differences.

[0147] This method can identify power stations whose performance is on the edge. For example, power station A09 has a final rating of "strong" with a maximum membership degree of 0.52, but its membership degree for the "moderate" rating is also 0.31. This membership degree distribution provides grid dispatchers with richer decision-making information, indicating that although the power station's performance is biased towards "strong," it also has some characteristics of the "moderate" rating.

[0148] In this embodiment, the fuzzy warning threshold τ = 0.5 is set. As can be seen from the results in Table 2, the maximum membership degree of all stations is not less than 0.5, therefore the fuzzy warning mechanism is not triggered.

[0149] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for evaluating the voltage support performance of new energy power stations based on dynamic response and fuzzy C-means clustering, characterized in that, include: Acquire data on several evaluation indicators for new energy power plants, including static potential indicators and dynamic performance indicators. The evaluation index data are combined and weighted to obtain the comprehensive weight vector of each index; The comprehensive weight vector is input into the fuzzy C-means clustering (FCM) model for iterative calculation to obtain the membership matrix of each station belonging to different performance levels. Based on the membership matrix, the output voltage support performance evaluation results are obtained. The evaluation index data are combined and weighted to obtain the comprehensive weight of each index, including: The subjective weight vector is calculated using the analytic hierarchy process (AHP). The objective weight vector is calculated using the entropy weight method; With the objective of minimizing the sum of squared deviations between the combined weights and the subjective and objective weights, an optimization model is established and solved to obtain the comprehensive weights. The subjective weight vector is calculated using the analytic hierarchy process (AHP) including: Based on pairwise importance comparisons of various indicators by several power system experts, a judgment matrix is ​​constructed. The eigenvector corresponding to the largest eigenvalue of the judgment matrix is ​​calculated and normalized to obtain the subjective weight vector. ; The objective weight vector is calculated using the entropy weight method, including: The original evaluation index data were dimensionless, and the entropy value of each index was calculated. and information utility value Thus, the objective weight vector is obtained. ; The optimization model is as follows: ， The constraints for optimizing the model are: ， In the formula, The combined weights to be determined; The subjective weights of each indicator; ; The comprehensive weights of each indicator are input into the fuzzy C-means clustering (FCM) model for iterative calculation to obtain the membership matrix of each station belonging to different performance levels, including: Initialize the number of clusters, fuzzy index, and convergence threshold; randomly initialize the membership matrix. Calculate the cluster center for each indicator data based on the current membership matrix and comprehensive weight vector; Iteratively update the cluster centers and membership matrix, calculate the change in the membership matrix, and stop iterating when the change in membership is less than a preset value.

2. The evaluation method according to claim 1, characterized in that, The static potential indicators include the capacity of new energy generating units, the capacity of static var generators (SVG), the reactive current ratio coefficient, the short-circuit ratio, and the impedance ratio. The dynamic performance indicators include reactive current response time and available energy storage capacity. The reactive current response time is the time required from the detection of a voltage drop at the grid connection point to the actual reactive current output by the new energy generating units and SVG reaching the preset command value ratio. The available energy storage capacity is the effective reactive power capacity that the energy storage system in the power station can provide during a fault.

3. The evaluation method according to claim 1, characterized in that, Initialize the number of clusters, fuzzy index, and convergence threshold; randomly initialize the membership matrix, including: Set the number of clusters The corresponding performance levels are "weak", "relatively weak", "average", "relatively strong", and "strong"; a fuzzy index is set. Set the convergence threshold Randomly initialize the membership matrix. ,in, The number of new energy power stations to be evaluated. For the first The station belongs to the first The initial membership degree of each cluster satisfies and ; Based on the current membership matrix and combined weight vector The cluster centers for each indicator data are calculated as follows: ， In the formula, As cluster center, For the first The 7-dimensional index vector of each station This indicates element-wise multiplication. For the first The station belongs to the first The membership degree of each cluster, For fuzzy index, This represents the number of new energy power stations to be evaluated.

4. The evaluation method according to claim 3, characterized in that, The membership matrix is ​​updated as follows: ， In the formula, For the first i The station belongs to the first k Update the membership degree of each cluster. For the number of clusters, , The first k The and the first j A new cluster center.

5. The evaluation method according to claim 4, characterized in that, The change in the membership matrix is ​​calculated as follows: ， In the formula, This represents the change in the membership matrix.

6. The evaluation method according to claim 1, characterized in that, The fuzzy index of the fuzzy C-means clustering FCM model Adaptive adjustment is performed based on the standard deviation of dynamic performance indicators, specifically: ， in, The standard deviation of dynamic performance indicators. The standard deviation is the static potential index.

7. The evaluation method according to claim 1, characterized in that, The voltage support performance evaluation results include: the final voltage support performance level, fuzzy information, and a fuzzy early warning mechanism; The final voltage support performance level is determined by the maximum membership degree, specifically: ， In the formula, For the final level of voltage support, To correspond to the performance level, Corresponding to "weak", Corresponding to "relatively weak" Corresponding to "generally" Corresponding to "relatively strong", Corresponding to "strong"; When the maximum membership value is lower than the preset threshold, a performance fuzzy warning is triggered.

Images