A performance evaluation method and system of an energy storage power station in a frequency modulation application scenario

Abstract

The application provides a performance evaluation method of an energy storage power station in a frequency modulation application scenario, comprising: obtaining monitoring data of the energy storage power station and power data participating in frequency modulation; performing correlation analysis and normalization processing on the monitoring data of the energy storage power station and the power data participating in frequency modulation and a pre-constructed evaluation index system to obtain positive indexes, inverse indexes and interval type indexes; setting weights of the positive indexes, the inverse indexes and the interval type indexes based on a data-driven weight optimization method to obtain a comprehensive score; and evaluating performance of the energy storage power station in the frequency modulation application scenario based on the comprehensive score; wherein the monitoring data comprises frequency modulation response conditions and PCS fault conditions; the characteristics and advantages of the energy storage power station in the frequency modulation application, such as fast speed, accurate control and bidirectional regulation capability, are fully considered, the frequency modulation performance of the energy storage power station is more accurately evaluated, and the energy storage power station promotes high-quality frequency modulation services for the system.

Description

Technical Field

[0001] This invention belongs to the field of power system demand response and frequency control technology, specifically relating to a performance evaluation method and system for energy storage power stations in frequency regulation application scenarios. Background Technology

[0002] Large-scale centralized grid connection of renewable energy generation has begun. However, the intermittency and randomness of renewable energy generation, coupled with load fluctuations, pose serious challenges to the power balance and frequency stability of the power grid. With the rapid development of new large-capacity energy storage technologies such as batteries, energy storage power station technology has begun to be applied in the frequency regulation field. Energy regulatory agencies in various provinces and cities have successively introduced relevant frequency regulation ancillary service market mechanisms to promote the application of energy storage in the power frequency regulation market. Compared with traditional generator frequency regulation methods, energy storage power stations have advantages such as high speed, precise control, and bidirectional regulation capabilities, and the frequency regulation quality they can provide is far superior to that of traditional generators.

[0003] Currently, we are in the early stages of power system reform, and there is a lack of experience in power market construction. A complete frequency regulation evaluation system that considers the advantages of energy storage power stations has not yet been established. Under the existing frequency regulation performance evaluation system, energy storage power stations and traditional generating units have the same evaluation effect when providing the same frequency regulation capacity. This clearly fails to reflect the technological advantages of energy storage power stations and is not conducive to the optimal allocation of frequency regulation resources. Therefore, establishing grid application performance evaluation indicators for energy storage power stations in frequency regulation scenarios based on their unique advantages is more urgent than ever. Summary of the Invention

[0004] To address the current shortcomings in the construction experience of the power market, the lack of a complete frequency regulation evaluation system that considers the advantages of energy storage power stations in terms of speed, precise control, and bidirectional regulation capabilities, and the resulting inefficiencies in the optimal allocation of frequency regulation resources, this invention provides a performance evaluation method for energy storage power stations in frequency regulation application scenarios, specifically including:

[0005] Acquire monitoring data from energy storage power stations and data on the amount of electricity used for frequency regulation;

[0006] Based on the monitoring data of the energy storage power station and the power data participating in frequency regulation, correlation analysis and normalization processing are performed with the pre-constructed evaluation index system to obtain positive indexes, inverse indexes and interval indexes.

[0007] A comprehensive score is obtained by setting weights for the positive, negative, and interval indicators based on a data-driven weight optimization method.

[0008] The performance of the energy storage power station in frequency regulation application scenarios is evaluated based on the comprehensive score.

[0009] The monitoring data includes: frequency modulation response status and PCS fault status.

[0010] Preferably, the construction of the evaluation index system includes:

[0011] Evaluation indicators for frequency regulation accuracy, frequency regulation correlation, frequency regulation response time, frequency regulation rate, SOC status of energy storage power station, and PCS failure rate of energy storage power station are determined based on the frequency regulation characteristics of energy storage power station.

[0012] The calculation formulas are used to quantify frequency regulation accuracy, frequency regulation correlation, frequency regulation response time, frequency regulation speed, SOC status of energy storage power station, and PCS failure rate of energy storage power station.

[0013] Preferably, the formula for calculating the frequency modulation speed is as follows:

[0014]

[0015] In the formula, S4 is the frequency modulation speed, P E P S The power output of the energy storage power station before and after the response, T E T S Set the start time and end time respectively.

[0016] Preferably, the formula for calculating the frequency modulation accuracy is as follows:

[0017]

[0018] In the formula, S1 is the difference between the frequency modulation demand and the frequency modulation mileage, and R t S t Let t be the frequency modulation mileage and frequency modulation demand at time t, respectively; V be the average frequency modulation demand within a scheduling cycle; and n be the number of frequency modulation signals within a scheduling cycle.

[0019] The formula for calculating the frequency modulation correlation is as follows:

[0020]

[0021] In the formula, S2 represents the correlation between the frequency modulation signal and the response value, and R... t+δ Let t+δ be the frequency regulation mileage at time t+δ, and δ be the response delay time of the energy storage power station;

[0022] The formula for calculating the frequency modulation response time is as follows:

[0023]

[0024] In the formula, S3 represents the time lag between the energy storage power station's response and the frequency modulation signal, and T... max This refers to the maximum response time of frequency modulation resources within a scheduling cycle;

[0025] The formula for calculating the SOC state of the energy storage power station is as follows:

[0026]

[0027] In the formula, S5 represents the SOC state of the energy storage power station, and C remain For the remaining electricity of the energy storage power station, C max This represents the maximum power capacity of the energy storage power station.

[0028] The formula for calculating the PCS failure rate of the energy storage power station is as follows:

[0029]

[0030] In the formula, S6 represents the PCS failure rate of the energy storage power station. Preferably, the positive, inverse, and interval-type indicators obtained by performing correlation analysis and normalization processing on the monitoring data and frequency regulation data of the energy storage power station with a pre-constructed evaluation index system include:

[0031] Based on the monitoring data of the energy storage power station and the power data participating in frequency regulation, the correlation coefficient, weight weakening factor, total adjustment factor of single indicator and indicator correlation adjustment factor of the evaluation indicators after the quantification process are calculated.

[0032] The evaluation indicators after correlation analysis are normalized to obtain positive indicators, inverse indicators, and interval indicators.

[0033] Preferably, the calculation of correlation coefficient, weight weakening factor, total adjustment factor for individual indicators, and correlation adjustment factor for the evaluation indicators after quantification includes:

[0034] The Pearson correlation coefficient was used to determine the correlation coefficient between the indicators;

[0035] Based on the calculation of the correlation coefficient, the weight of each indicator is weakened, and the weight weakening factor between indicators is calculated.

[0036] Calculate the total adjustment factor for a single indicator based on the weighting weakening factor among indicators;

[0037] Based on the total adjustment factor of the individual indicator, the correlation adjustment factor of the indicator is calculated.

[0038] Preferably, the normalization process of the evaluation indicators after correlation analysis to obtain positive indicators, inverse indicators, and interval indicators includes:

[0039] Based on the membership function fitted to the data distribution, the evaluation indicators after correlation analysis are divided into positive indicators, inverse indicators, and interval indicators.

[0040] Based on the membership function fitted by the data distribution, the membership function type of the evaluation index for each category is selected for homogenization.

[0041] A unified scoring standard is applied based on the membership function type selected from the evaluation indicators of each category.

[0042] The positive indicators include: frequency modulation accuracy, frequency modulation correlation, frequency modulation rate, and frequency modulation response time.

[0043] The negative indicators include: PCS failure rate of energy storage power stations;

[0044] The interval-type indicators include: the SOC status of the energy storage power station.

[0045] Preferably, the membership function based on data distribution fitting, which selects membership function types for evaluation indicators of each category to achieve convergence, includes:

[0046] For positive indices, the membership function of the upper type is used; for inverse indices, the membership function of the lower type is used; and for interval indices, the membership function of the middle type is used.

[0047] Preferably, the data-driven weight optimization method assigns weights to the positive indicators, inverse indicators, and interval indicators to obtain a comprehensive score, including:

[0048] Based on the data-driven weight optimization method, the positive index, the inverse index and the interval index are assigned weights to obtain a comprehensive score, and the objective weights are set using the entropy weight method and the coefficient of variation method.

[0049] A comprehensive score is obtained based on the positive indicators, inverse indicators, interval indicators, and weights.

[0050] Preferably, the data-driven weight optimization method, in which weights are assigned to the positive, negative, and interval indicators to obtain a comprehensive score, and objective weights are set using the entropy weight method and the coefficient of variation method, includes:

[0051] Based on the entropy weight method and the coefficient of variation method, considering the information content and data distribution of the positive index, the inverse index and the interval index, the initial objective weights are obtained.

[0052] Based on the initial objective weights and the correlation adjustment factor, a comprehensive weight is obtained.

[0053] Preferably, the comprehensive score is calculated using the following formula:

[0054]

[0055] In the formula, F kFor the score of the k-th energy storage power station, S i The quantification result of the i-th indicator, where n is the total number of energy storage power stations. This is the overall weighting coefficient.

[0056] Based on the same concept, the present invention provides a performance evaluation system for energy storage power stations in frequency regulation application scenarios, including: an acquisition module, a processing module, a scoring module and an evaluation module;

[0057] The acquisition module is used to acquire monitoring data of the energy storage power station and power data participating in frequency regulation;

[0058] The processing module performs correlation analysis and normalization on the monitoring data of the energy storage power station and the power data participating in frequency regulation with the pre-constructed evaluation index system to obtain positive indicators, inverse indicators and interval indicators.

[0059] The scoring module is used to set weights for the positive indicators, inverse indicators, and interval indicators based on a data-driven weight optimization method to obtain a comprehensive score.

[0060] The evaluation module is used to evaluate the performance of the energy storage power station in frequency regulation application scenarios based on the comprehensive score;

[0061] The monitoring data includes: frequency modulation response status and PCS fault status.

[0062] Preferably, the processing module includes: an indicator submodule and a quantification submodule;

[0063] The indicator submodule is used to determine evaluation indicators such as frequency regulation accuracy, frequency regulation correlation, frequency regulation response time, frequency regulation rate, SOC status of energy storage power station, and PCS failure rate of energy storage power station based on the frequency regulation characteristics of energy storage power station.

[0064] The quantization submodule performs quantification based on the calculation formulas for frequency modulation accuracy, frequency modulation correlation, frequency modulation response time, frequency modulation speed, SOC status of the energy storage power station, and PCS failure rate of the energy storage power station.

[0065] Preferably, the processing module further includes: an analysis submodule and a normalization submodule;

[0066] The analysis submodule calculates the correlation coefficient, weight weakening factor, total adjustment factor for single indicator, and indicator correlation adjustment factor for the evaluation indicators after quantitative processing, based on the monitoring data of the energy storage power station and the power data participating in frequency regulation.

[0067] The normalization submodule is used to normalize the evaluation indicators after correlation analysis to obtain positive indicators, inverse indicators, and interval indicators.

[0068] Preferably, the scoring module includes: a weight setting submodule and a comprehensive scoring submodule;

[0069] The weight setting submodule is used for a data-driven weight optimization method. It sets weights for the positive, negative and interval indicators to obtain a comprehensive score, and uses the entropy weight method and the coefficient of variation method to set objective weights.

[0070] The comprehensive scoring submodule is used to obtain a comprehensive score based on the positive indicators, inverse indicators, interval indicators, and weights.

[0071] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0072] 1. This invention provides a performance evaluation method for energy storage power stations in frequency regulation application scenarios, comprising: acquiring monitoring data of the energy storage power station and power data participating in frequency regulation; performing correlation analysis and normalization processing on the monitoring data of the energy storage power station and the power data participating in frequency regulation with a pre-constructed evaluation index system to obtain positive indicators, inverse indicators, and interval indicators; setting weights for the positive indicators, inverse indicators, and interval indicators based on a data-driven weight optimization method to obtain a comprehensive score; evaluating the performance of the energy storage power station in frequency regulation application scenarios based on the comprehensive score; wherein, the monitoring data includes: frequency regulation response status and PCS fault status; fully considering the advantages of energy storage power stations in frequency regulation applications, such as high speed, precise control, and bidirectional adjustment capabilities, to more accurately evaluate the frequency regulation performance of energy storage power stations and promote energy storage power stations to bring high-quality frequency regulation services to the system. Attached Figure Description

[0073] Figure 1 This invention provides a method flowchart for evaluating the performance of an energy storage power station in a frequency regulation application scenario;

[0074] Figure 2 A flowchart of an evaluation system for an energy storage power station in a frequency regulation application scenario, provided by an embodiment of the present invention;

[0075] Figure 3 A detailed algorithm diagram for comprehensive scoring is provided in an embodiment of the present invention. Detailed Implementation

[0076] The embodiments of the present invention will be further described with reference to the accompanying drawings.

[0077] Example 1:

[0078] This invention provides a performance evaluation method for energy storage power stations in frequency regulation application scenarios, combined with Figure 1 The method flowchart is introduced, specifically including:

[0079] Step 1: Obtain monitoring data of the energy storage power station and power data participating in frequency regulation;

[0080] Step 2: Based on the monitoring data of the energy storage power station and the power data participating in frequency regulation, perform correlation analysis and normalization processing with the pre-constructed evaluation index system to obtain positive indexes, inverse indexes, and interval indexes;

[0081] Step 3: Using a data-driven weight optimization method, assign weights to the positive, negative, and interval indicators to obtain a comprehensive score;

[0082] Step 4: Evaluate the performance of the energy storage power station in frequency regulation application scenarios based on the comprehensive score;

[0083] Step 1 involves acquiring monitoring data from the energy storage power station and the electricity data participating in frequency regulation, specifically including:

[0084] 1. Compared with traditional generator frequency regulation, energy storage power stations have the advantages of fast frequency regulation speed, precise control and bidirectional regulation. In order to fully reflect the advantages of energy storage power stations in frequency regulation and promote the optimal allocation of energy storage power resources;

[0085] Step 2: Based on the monitoring data of the energy storage power station and the power data participating in frequency regulation, correlation analysis and normalization processing are performed with the pre-constructed evaluation index system to obtain positive indicators, inverse indicators, and interval indicators, specifically including:

[0086] 1) Considering the advantages of energy storage power stations participating in frequency regulation, determine more comprehensive evaluation indicators;

[0087] 2) Quantify each evaluation indicator and perform correlation analysis;

[0088] 3) In order to unify the basic measurement units of different indicators, the indicators are normalized.

[0089] 2. Frequency regulation indicators, mainly including: frequency regulation accuracy, frequency regulation correlation, frequency regulation response time, frequency regulation rate, SOC status, and PCS failure rate of energy storage power station.

[0090] 3. New performance metrics are introduced, including frequency modulation accuracy, frequency modulation correlation, frequency modulation response time, frequency modulation rate, SOC status, and PCS failure rate. These metrics are quantified as follows:

[0091] (1) Frequency modulation accuracy: the difference between frequency modulation demand and frequency modulation mileage.

[0092] Among them, R t S tLet t represent the frequency modulation mileage and frequency modulation demand at time t, respectively; V represent the average frequency modulation demand within a scheduling cycle; and n represent the number of frequency modulation signals within a scheduling cycle.

[0093] (2) Frequency modulation correlation: The degree of correlation between the frequency modulation signal and the response value;

[0094]

[0095] Among them, R t+δ Let t+δ be the frequency regulation mileage at time t+δ, where δ is the response delay time of the energy storage power station, and n is the number of frequency regulation signals in one scheduling cycle;

[0096] (3) Frequency modulation response time: The time by which the energy storage power station's response lags behind the frequency modulation signal.

[0097] Among them, T max The maximum response time of frequency regulation resources within a scheduling cycle is δ, where δ is the response delay time of the energy storage power station.

[0098] (4) Frequency regulation speed: the ramp-up speed of the energy storage power station

[0099] Among them, P E P S The power output of the energy storage power station before and after the response, T E T S Set the start time and end time respectively;

[0100] (5) Energy storage power station SOC:

[0101] Among them, C remain For the remaining electricity of the energy storage power station, C max This represents the maximum power capacity of the energy storage power station.

[0102] (6) PCS failure rate of energy storage power station:

[0103] 4. To avoid double counting of indicators, 2) includes:

[0104] 2.1. The Pearson correlation coefficient is used to determine the correlation coefficient between each pair of indicators;

[0105]

[0106] in, x ij Let j be the value of the i-th indicator in the j-th group. Then the average value of each indicator, Let be the standard deviation of the i-th indicator. Let x be the standard deviation of the j-th indicator.ik Let x be the i-th indicator of the k-th energy storage power station. jk Let X be the i-th indicator of the k-th energy storage power station. i Let X be the i-th indicator. j Let j be the j-th indicator.

[0107] 2.2 Calculate the weight weakening factor between pairs of indicators.

[0108] The correlation coefficient reflects the consistency of trends between two indicators. A larger correlation coefficient indicates a more pronounced correlation between the trends of the two indicators, but its actual representativeness is lower. Based on the calculation of the correlation coefficient, the weight of each indicator is weakened, and the actual representativeness of each indicator is (1-0.5|r ij |), that is, the weight weakening factor between pairs of indicators is:

[0109]

[0110] 2.3 Calculate the total adjustment factor for individual indicators

[0111] For any given indicator, it may be correlated with multiple other indicators. Therefore, the total adjustment factor should be the product of its pairwise adjustment factors with all other indicators.

[0112]

[0113] 2.4 Calculate the correlation adjustment factor for the indicators

[0114] The total adjustment factor exhibits a significant bipolar distribution, which may amplify or diminish the effects of certain indicators. Therefore, it needs to be smoothed to avoid excessively extreme distribution. Log-normalizing the adjustment factor yields the correlation adjustment factor.

[0115]

[0116] 5. To summarize and unify the statistical distribution of various quantitative indicators and standardize the basic units of measurement, the scores of each indicator need to be normalized. Although the indicator data of the energy storage power station performance evaluation system are relatively specific and complete, the performance of each energy storage power station is inherently ambiguous. Therefore, a membership function based on data distribution fitting is proposed to normalize the scores.

[0117] First, the nature of the indicators needs to be classified. For the evaluation system of energy storage power stations, they can be divided into positive indicators, inverse indicators, and intermediate indicators. We should first homogenize them, mapping the value of each indicator to a specific score under a unified standard, so that they are comparable to each other.

[0118] Based on the data distribution characteristics of each indicator, the membership function for each indicator is determined. The membership function can be linear or non-linear, depending on the distribution of the indicator data. There are three main types of membership function distributions: positive, negative, and intermediate. For positive indicators, a positive membership function is used; for negative indicators, a negative membership function is used; and for interval indicators, an intermediate membership function is used.

[0119] Step 3: Based on a data-driven weight optimization method, weights are assigned to the positive indicators, inverse indicators, and interval indicators to obtain a comprehensive score, specifically including:

[0120] 4. A data-driven weight optimization method is used to set weights for each indicator;

[0121] 5. Establish a multi-dimensional evaluation index system for energy storage power stations in frequency regulation applications;

[0122] 6. Classify the nature of the indicators, homogenize different types of indicators, and map the value of each indicator to a specific score under a unified standard to make them comparable. This can be done by classifying the indicators as follows:

[0123] Positive indicators: The higher the original value of an indicator, the higher the corresponding score should be. In this evaluation system, frequency modulation accuracy, frequency modulation correlation, frequency modulation rate, and frequency modulation response time are all positive indicators.

[0124] Inverse indicator: The higher the original value of the indicator, the lower the corresponding score should be. For example, the PCS failure rate of an energy storage power station.

[0125] Range-type indicators: The highest score is achieved when the indicator value falls within a certain range. For both upward and downward frequency regulation conditions of energy storage power stations, the frequency regulation performance is most balanced when the SOC of the energy storage power station is between 0.4 and 0.6, and thus a higher score should be obtained.

[0126] 7. A data-driven weight optimization method is used to set the weights of each performance metric.

[0127] To address the issue that the current performance evaluation weighting of energy storage power stations is heavily influenced by subjective experience, and considering the characteristics of energy storage power station performance index data being disorganized, having large ranges, and being complete, this paper adopts the entropy weighting method and the coefficient of variation method to set objective weights from a data-driven perspective.

[0128] Taking into account both entropy weight and coefficient of variation, since these two methods consider information content and data distribution respectively and have no clear primary or secondary relationship, the initial objective weight is obtained by directly averaging them.

[0129]

[0130] In the formula, m represents the number of energy storage power station indicators;

[0131] Based on the analysis of the correlation adjustment factor and the initial objective weights, the comprehensive weights are obtained and then normalized:

[0132]

[0133] in, G is the comprehensive weighting coefficient. i is the normalized correlation adjustment factor, and m is the number of performance indicators of the energy storage power station;

[0134] 8. From a data-driven perspective, use the entropy weight method and the coefficient of variation method to set objective weights.

[0135] (1) Entropy weight method

[0136] Entropy weighting is an objective weighting method that applies information entropy theory to determine objective weights, reflecting the amount of information presented in the original data of each indicator. If the numerical difference of an indicator is large, its entropy is small, reflecting more information, and its corresponding weight is also larger. We compared and analyzed the entropy weighting method, the analytic hierarchy process (AHP), and the Delphi method. When the data is relatively complete, the entropy weighting method has the highest accuracy in setting weights. For energy storage power stations, the selected indicator values ​​are easy to obtain and the data is complete. However, due to the scale and characteristic parameters of different power stations, there may be significant differences between the indicator values. Therefore, the entropy weighting method is suitable for setting weights.

[0137] For each set of data containing values ​​for various indicators, calculate the proportion of the i-th indicator in the j-th set of data.

[0138]

[0139] Calculate the entropy value of the i-th index.

[0140] Where e i All are positive values.

[0141] Then the entropy weight of the i-th index is

[0142] In the formula, e i Let e ​​be the entropy value of the i-th index. i It is a positive value;

[0143] (2) Coefficient of variation method

[0144] The coefficient of variation reflects the degree of difference between indicators and is suitable for situations where the indicators are relatively independent. The wider the distribution range of the indicators, the larger the coefficient of variation. For energy storage power stations, the data of each indicator are relatively independent. At the same time, the entropy weight is greatly affected by special data. In order to reduce the impact of the special operating conditions of energy storage power stations on the evaluation system, we use the coefficient of variation method to balance the shortcomings of the entropy weight.

[0145] The coefficient of variation of the i-th indicator is

[0146] Where, σ i The standard deviation of the indicator. This represents the average value of the indicators.

[0147] The weights of the coefficients of variation for each indicator are then:

[0148] Among them, V i Let be the coefficient of variation of the i-th indicator.

[0149] 9. Based on the quantified results of the aforementioned indicators and the weighting settings, a comprehensive score can be obtained by weighting the results.

[0150]

[0151] Among them, F k For the score of the k-th energy storage power station, S i The quantification result of the i-th indicator, where n is the total number of energy storage power stations and m is the number of indicators for energy storage power stations. This is the overall weighting coefficient.

[0152] Step 4: Evaluate the performance of the energy storage power station in frequency regulation application scenarios based on the comprehensive score, specifically including:

[0153] To fully demonstrate the advantages of energy storage power stations in frequency regulation, the frequency regulation performance is no longer measured by a single frequency regulation capacity, but by selecting more performance indicators to evaluate the frequency regulation performance of energy storage power stations.

[0154] Example 2:

[0155] Based on the same concept, this invention provides a performance evaluation system for energy storage power stations in frequency regulation application scenarios, such as... Figure 2 As shown, it specifically includes:

[0156] Step 1: The module acquires monitoring data from the energy storage power station and the power data participating in frequency regulation;

[0157] Step 2: The processing module performs correlation analysis and normalization on the monitoring data of the energy storage power station and the power data participating in frequency regulation with the pre-constructed evaluation index system to obtain positive indexes, inverse indexes and interval indexes;

[0158] Step 3: The scoring module uses a data-driven weight optimization method to assign weights to the positive indicators, inverse indicators, and interval indicators to obtain a comprehensive score;

[0159] Step 4: The evaluation module evaluates the performance of the energy storage power station in frequency regulation application scenarios based on the comprehensive score.

[0160] Step 1: The acquisition module acquires monitoring data from the energy storage power station and power data participating in frequency regulation, specifically including:

[0161] Compared to traditional generator frequency regulation, energy storage power stations have faster frequency regulation speed, more precise control, and bidirectional adjustment capabilities. In order to fully reflect the advantages of energy storage power stations in frequency regulation and promote the optimal allocation of energy storage power resources;

[0162] Step 2: The processing module performs correlation analysis and normalization on the monitoring data of the energy storage power station and the power data participating in frequency regulation with the pre-constructed evaluation index system to obtain positive indicators, inverse indicators, and interval indicators, specifically including:

[0163] Evaluation metrics include: frequency modulation accuracy, frequency modulation correlation, frequency modulation response time, frequency modulation rate, and state of charge (SOC).

[0164] 2.1 Because there may be some correlation between the selected indicators, such as frequency modulation correlation and frequency modulation accuracy, which will affect the accuracy of the evaluation, the Pearson correlation coefficient is used to determine the correlation coefficient between each pair of indicators.

[0165]

[0166] in,

[0167] 2.2 Calculate the weight weakening factor between pairs of indicators.

[0168] The correlation coefficient reflects the consistency of trends between two indicators. The larger the correlation coefficient, the more obvious the correlation between the trends of the two indicators, and the less representative they are.

[0169] If the correlation coefficient between two indicators is not zero, then the actual representativeness of each indicator is (1-0.5|r ij Therefore, the weight of each indicator should be weakened, that is, the weight weakening factor between each pair of indicators is:

[0170] 2.3 Calculate the total adjustment factor for individual indicators

[0171] For any given indicator, it may be correlated with multiple other indicators; therefore, the total adjustment factor should be the product of its pairwise adjustment factors with all other indicators.

[0172] 2.4 Calculate the correlation adjustment factor for the indicators

[0173] The total adjustment factor exhibits a severe polarization, which may amplify or diminish the effect of certain indicators. Therefore, it needs to be smoothed to avoid an overly extreme distribution.

[0174] Log-normalizing the adjustment factor yields the correlation adjustment factor as follows:

[0175] To standardize the statistical distribution of various quantitative indicators and ensure a consistent basic unit of measurement, the scores of each indicator need to be normalized. Although the performance evaluation system for energy storage power stations has relatively specific and complete indicator data, the performance of each energy storage power station is inherently ambiguous. Therefore, a membership function based on data distribution fitting is proposed to normalize the scores.

[0176] First, it is necessary to classify the nature of the indicators.

[0177] Because positive indicators, inverse indicators, and interval indicators cannot be directly used for weighted calculations, they must first be homogenized, mapping the value of each indicator to a specific score under a unified standard to make them comparable to each other.

[0178] Based on the data distribution characteristics of each indicator, the membership function of each indicator is determined.

[0179] Membership functions can be linear or non-linear, and the appropriate fit depends on the distribution of the index data.

[0180] There are three main types of distributions of conventional membership functions: the upper type, the lower type, and the intermediate type.

[0181] For positive indices, a membership function of the Rong-type is used;

[0182] For inverse indices, a membership function of the Rong-type is used;

[0183] For interval indicators, an intermediate membership function is used.

[0184] By classifying the nature of indicators, homogenizing different types of indicators, and mapping the value of each indicator to a specific score under a unified standard to make them comparable, they can be categorized as follows:

[0185] Positive indicators: The larger the original value of an indicator, the higher the corresponding score should be.

[0186] In this evaluation system, frequency regulation accuracy, frequency regulation correlation, frequency regulation rate, and the operational availability of PCS for energy storage power stations are all positive indicators.

[0187] Inverse indicators: The larger the original value of an indicator, the lower the corresponding score should be, such as frequency modulation response time.

[0188] Interval-based indicators: When the original value of an indicator is within a certain fixed range, it will receive a higher score, such as the SOC status.

[0189] Step 3: The scoring module uses a data-driven weight optimization method to assign weights to the positive, negative, and interval indicators to obtain a comprehensive score, specifically including:

[0190] A data-driven weight optimization method is used to set the weights for each performance metric.

[0191] To address the issue that the current performance evaluation weighting of energy storage power stations is heavily influenced by subjective experience, and considering the chaotic distribution and large range of performance index data for energy storage power stations, this paper adopts the entropy weighting method and the coefficient of variation method to set objective weights from a data-driven perspective.

[0192] like Figure 3 The method for setting the weights of this system comprehensively considers entropy weights and the coefficient of variation. Since these two methods consider information content and data distribution respectively, and there is no obvious primary or secondary relationship, the initial objective weights are obtained by directly averaging them:

[0193]

[0194] Based on the analysis of the correlation adjustment factor and the initial objective weights, the comprehensive weights are obtained and then normalized:

[0195]

[0196] From a data-driven perspective, the entropy weight method and the coefficient of variation method are used to set objective weights.

[0197] Entropy weighting is an objective weighting method that uses information entropy theory to determine objective weights, reflecting the amount of information presented in the original data of each indicator.

[0198] If a certain indicator has a large numerical difference, its entropy is small, it reflects more information, and its corresponding weight is also larger.

[0199] For energy storage power stations, the selected indicator values ​​are easy to obtain and the data is complete. However, due to the scale and characteristic parameters of each power station, there may be significant differences between the indicator values. Therefore, the entropy weight method is suitable for setting the weights.

[0200] For each set of data containing values ​​for various indicators, calculate the proportion of the i-th indicator in the j-th set of data.

[0201] Calculate the entropy value of the i-th index.

[0202] Where e i All are positive values.

[0203] Then the entropy weight of the i-th index is

[0204] The coefficient of variation reflects the degree of dispersion of the index distribution.

[0205] The wider the distribution range of the indicators, the greater the coefficient of variation. Entropy weight is greatly affected by special data. The coefficient of variation can balance the defects of entropy weight to a certain extent.

[0206] The coefficient of variation of the i-th indicator is

[0207] Where, σ i The standard deviation of the indicator. This represents the average value of the indicators.

[0208] The weights of the coefficients of variation for each indicator are then:

[0209] Finally, based on the quantified results of the aforementioned indicators and the weighted results, a comprehensive score can be obtained by weighting the results.

[0210]

[0211] In the formula, F k The score for the k-th energy storage power station.

[0212] Step 4: The evaluation module evaluates the performance of the energy storage power station in frequency regulation application scenarios based on the comprehensive score, specifically including:

[0213] Compared to traditional generator frequency regulation, energy storage power stations have faster frequency regulation speed, more precise control, and bidirectional adjustment capabilities. To fully reflect the advantages of energy storage power stations in frequency regulation, the single frequency regulation capacity is no longer used to measure frequency regulation performance. Instead, more performance indicators are selected to evaluate the frequency regulation performance of energy storage power stations.

[0214] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0215] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0216] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0217] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0218] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, 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 method for evaluating performance of an energy storage power station in a frequency modulation application scenario, characterized in that, The application comprises the following steps: obtaining monitoring data of the energy storage power station and power data participating in frequency modulation; performing correlation analysis and normalization processing on the monitoring data of the energy storage power station and the power data participating in frequency modulation based on a pre-constructed evaluation index system to obtain positive indexes, inverse indexes and interval type indexes; setting weights for the positive indexes, inverse indexes and interval type indexes based on a data-driven weight optimization method to obtain a comprehensive score; evaluating the performance of the energy storage power station in the frequency modulation application scenario based on the comprehensive score; wherein the monitoring data comprises frequency modulation response and PCS failure conditions; the construction of the evaluation index system comprises the following steps: determining evaluation indexes of frequency modulation accuracy, frequency modulation correlation, frequency modulation response time, frequency modulation speed, energy storage power station SOC state and energy storage power station PCS failure rate based on the frequency modulation characteristics of the energy storage power station; and performing quantitative processing based on the calculation formulas of frequency modulation accuracy, frequency modulation correlation, frequency modulation response time, frequency modulation speed, energy storage power station SOC state and energy storage power station PCS failure rate; the correlation analysis and normalization processing on the monitoring data of the energy storage power station and the power data participating in frequency modulation based on the pre-constructed evaluation index system to obtain positive indexes, inverse indexes and interval type indexes comprises the following steps: calculating correlation coefficients, weight weakening factors, single index total adjustment factors and index correlation adjustment factors of the evaluation indexes subjected to the quantitative processing based on the monitoring data of the energy storage power station and the power data participating in frequency modulation; performing normalization processing on the evaluation indexes subjected to the correlation analysis to obtain positive indexes, inverse indexes and interval type indexes; the normalization processing on the evaluation indexes subjected to the correlation analysis to obtain positive indexes, inverse indexes and interval type indexes comprises the following steps: dividing the evaluation indexes subjected to the correlation analysis into positive indexes, inverse indexes and interval type indexes based on a membership function of data distribution fitting; selecting a membership function type for each classification of the evaluation indexes subjected to the correlation analysis to perform homotopy; selecting a membership function type for each classification of the evaluation indexes to perform a unified scoring standard; wherein the positive indexes comprise frequency modulation accuracy, frequency modulation correlation, frequency modulation speed and frequency modulation response time; the inverse indexes comprise energy storage power station PCS failure rate; the interval type indexes comprise energy storage power station SOC state.

2. The method of claim 1, wherein, The calculation formula of the frequency modulation speed is as follows: In the formula, is the frequency modulation speed, , respectively, the output of the energy storage power station before and after the response, , respectively, the start time and the end time.

3. The method of claim 1, wherein, The calculation formula of the frequency modulation accuracy is as follows: In the formula, is the difference between the frequency modulation demand and the frequency modulation mileage, , respectively, the frequency modulation mileage and the frequency modulation demand at the moment, is the average of the frequency modulation demand in a dispatch cycle, is the number of frequency modulation signals in a dispatch cycle; The calculation formula of the frequency modulation correlation is as follows: wherein is the degree of correlation between the frequency modulation signal and the response value, is the is the frequency modulation mileage at the time instant, is the energy storage plant response delay time; The calculation formula of the frequency modulation response time is as follows: Equation, the time for the energy storage power station to respond to the frequency modulation signal, the maximum response time of the frequency modulation resource in a scheduling period; The calculation formula of the energy storage power station SOC state is as follows: wherein SOC is the state of charge of the energy storage plant, R is the remaining amount of electricity of the energy storage plant, M is the maximum amount of electricity of the energy storage plant; The calculation formula of the energy storage power station PCS failure rate is as follows: In the formula, is the PCS failure rate of the energy storage power station.

4. The method of claim 1, wherein, the calculation of correlation coefficients, weight weakening factors, single index total adjustment factors and index correlation adjustment factors of the evaluation indexes subjected to the quantitative processing comprises the following steps: determining the correlation coefficients between indexes by using Pearson correlation coefficients; weakening the weight of each index based on the calculation of the correlation coefficients to calculate the weight weakening factors between indexes; calculating the single index total adjustment factors based on the weight weakening factors between indexes; calculating the index correlation adjustment factors based on the single index total adjustment factors.

5. The method of claim 1, wherein, The membership function based on data distribution fitting is same trend for each classification evaluation index selection membership function type, including: For positive indicators, adopt the upper membership function; for inverse indicators, adopt the lower membership function; for interval type indicators, adopt the intermediate type membership function.

6. The method of claim 4, wherein, The weight optimization method based on data driving sets weights for the positive indicators, inverse indicators and interval type indicators to obtain comprehensive scores, including: The weight optimization method based on data driving sets weights for the positive indicators, inverse indicators and interval type indicators to obtain comprehensive scores, adopts entropy weight method and coefficient of variation method to set objective weights; According to the positive indicators, inverse indicators and interval type indicators and the weights, comprehensive scores are obtained.

7. The method of claim 6, wherein, The weight optimization method based on data driving sets weights for the positive indicators, inverse indicators and interval type indicators to obtain comprehensive scores, adopts entropy weight method and coefficient of variation method to set objective weights, including: Based on the entropy weight method and the coefficient of variation method, considering the information amount and data distribution of the positive indicators, inverse indicators and interval type indicators, initial objective weights are obtained; Based on the initial objective weights and the correlation adjustment factor, comprehensive weights are obtained.

8. The method of claim 6, wherein, The calculation formula of the comprehensive score is as follows: In the formula, is the score of the i-th energy storage power station, is the quantitative result of the i-th index, is the total number of energy storage power stations, is the comprehensive weight coefficient.​ 9. A performance evaluation system of energy storage plants in frequency regulation application scenarios, for the method of any of claims 1-8, characterized in that, Including: An acquisition module, a processing module, a scoring module and an evaluation module; The acquisition module is used to acquire monitoring data of the energy storage power station and power consumption data participating in frequency modulation; The processing module performs correlation analysis and normalization processing on the monitoring data of the energy storage power station and the power consumption data participating in frequency modulation based on a pre-constructed evaluation index system to obtain positive indicators, inverse indicators and interval type indicators; The scoring module is used to set weights for the positive indicators, inverse indicators and interval type indicators based on the weight optimization method based on data driving to obtain comprehensive scores; The evaluation module is used to evaluate the performance of the energy storage power station in the frequency modulation application scenario based on the comprehensive scores; The monitoring data includes frequency modulation response and PCS failure conditions. Including: an index submodule and a quantization submodule; The index submodule is used to determine evaluation indexes of frequency modulation accuracy, frequency modulation correlation, frequency modulation response time, frequency modulation speed, energy storage power station SOC state and energy storage power station PCS failure rate based on the frequency modulation characteristics of the energy storage power station; The quantization submodule performs quantization processing based on the calculation formula of the frequency modulation accuracy, frequency modulation correlation, frequency modulation response time, frequency modulation speed, energy storage power station SOC state and energy storage power station PCS failure rate.

10. The system of claim 9, wherein, The processing module further includes an analysis submodule and a normalization submodule; The analysis submodule calculates correlation coefficients, weight weakening factors, single indicator total adjustment factors and index correlation adjustment factors based on the monitoring data of the energy storage power station and the power consumption data participating in frequency modulation for the evaluation indexes subjected to the quantization processing; The normalization submodule is used to perform normalization processing on the evaluation indexes subjected to the correlation analysis to obtain positive indicators, inverse indicators and interval type indicators.

11. The system of claim 10, wherein, The scoring module includes a weight setting submodule and a comprehensive score submodule; The weight setting submodule is configured to set weights of the positive indicators, the inverse indicators and the interval type indicators based on a data-driven weight optimization method to obtain a comprehensive score, and set objective weights by using an entropy weight method and a coefficient of variation method. The comprehensive score submodule is configured to obtain the comprehensive score according to the positive indicators, the inverse indicators and the interval type indicators and the weights.

Images