Correlation analysis-based grid-connected wind power plant operation performance evaluation method and system

Abstract

The invention discloses a grid-connected wind power plant operation performance evaluation method and system based on correlation analysis, and the method comprises the steps: collecting a grid-connected performance evaluation index evaluation data source, and forming historical sampling data; a test data set is obtained after preprocessing; performing correlation analysis on the test data set, selecting feature column data with high correlation degree to form a new test data set, and constructing a wind power plant operation performance evaluation model; mining the correlation between a preset grid-connected wind power plant operation performance evaluation index and each influence factor in the new test data set, and converting the correlation into a quantitative evaluation index; substituting the actual operation data of the wind power plant into the model, calculating the numerical value of each performance index of the wind power plant, and evaluating the performance of each performance index of the wind power plant in combination with the quantitative evaluation index to obtain the operation performance evaluation result of the grid-connected wind power plant. According to the method, the problem that the index coverage is small in an existing evaluation method is solved, and the operation performance index evaluation system of the wind turbine generator can be established more comprehensively and deeply.

Description

Technical Field

[0001] The present invention belongs to the field of electric power systems, and in particular relates to a method and system for evaluating the operating performance of a grid-connected wind farm based on correlation analysis. Background Art

[0002] With the rapid development of wind power technology and the gradual increase in wind power penetration, the impact of grid-connected wind power on the stability of the power system is also increasing, and the operating performance and production efficiency of wind turbines have also become key factors affecting the economic benefits of wind farms. However, due to the complex and changeable operating environment of wind turbines and the mutual influence of internal parameters of the units, the volatility and randomness of wind power generation have brought a series of problems that need to be solved after it is connected to the grid. Focusing on the evaluation of wind turbine operating performance, fully exploring the potential data associations between wind energy resource data, unit power generation data and grid-connected operation data, conducting a more comprehensive analysis and evaluation of the operating performance of wind farms, and establishing an indicator evaluation system covering the power generation characteristics of all links of wind power production are effective ways to efficiently utilize wind power resources, and can also provide good technical support for the grid to absorb wind power resources on a larger scale.

[0003] Document 1 "Construction and Application of Evaluation Index System for Large-Scale Wind Power Grid Connection" (Power System Technology, Vol. 45, No. 3, 2021, p. 841) established an evaluation index system based on the "five rates" of wind power installed capacity penetration rate, power penetration rate, electricity penetration rate, utilization rate and power abandonment rate from the perspective of large-scale wind turbine grid connection and absorption. It evaluated the impact of different utilization rate indicators and restrictions on wind power installed capacity penetration rate, and used actual grid operation data to illustrate the characteristics of wind power generation and the application method of the "five rates" index system, which has certain guiding significance for wind power planning and improving the lean operation level of wind farms.

[0004] Literature 1 focuses on the grid-connected operation of wind turbines. The classification of indicator categories in the performance evaluation process is clearer, but the formulation of the "five rates" indicators only focuses on the grid side of the grid-connected operation of wind turbines. It is unable to evaluate the operating characteristics of the "power supply side" of wind turbines and the utilization of wind energy resources by the turbines. It can only supplement and improve the "grid side" indicators of wind turbines participating in the grid-connected system operation.

[0005] Literature 2, "Multi-temporal and spatial scale wind power statistical characteristics evaluation index system and its application" (Proceedings of the CSEE, Vol. 33, No. 13, 2013, p. 53), divides the performance evaluation indicators of wind power generation into two categories: indicators that reflect the natural characteristics of wind power based on wind resources and wind power characteristics themselves, and indicators that reflect the interactive characteristics of the power grid that require comprehensive data such as power grids and loads to obtain. This has enriched the performance evaluation index volume of wind farm operations to a certain extent and provided more statistical indicator references for research work.

[0006] Literature 2 takes the characteristics of wind resources / wind power itself and the physical interaction between wind power, power grid and load as the starting point, and seeks expanded indicators for performance evaluation in the wind resource absorption link and power grid consumption link, which improves the evaluation indicator resource library to a certain extent. However, the evaluation link of this literature starts from the perspective of the physical meaning of performance indicators, and it is difficult to explore the operating indicators that cannot determine the correlation, lack qualitative analysis or quantitative description, and cannot conduct a more in-depth and extensive exploration of wind power.

[0007] Literature 3 "Application of the Analytic Hierarchy Process in Economic Evaluation of Wind Farm Operation" (China Electric Power, Vol. 39, No. 9, 2006, p. 42) uses the analytic hierarchy process to classify the factors that affect the economic operation of wind farms. From the perspective of wind farm operation management, it evaluates the economic aspects of wind farm operation such as power generation, operating costs and personnel efficiency through pairwise comparison. The overall ranking of the relative importance of each influencing factor is then determined based on the work experience of the scorers.

[0008] Literature 3 evaluates multiple indicators as a whole by classifying and interpreting the sampled data, and attempts to evaluate the economic efficiency of each wind farm operation from the perspective of evaluation methods. However, when there are many factors such as uncertainty and fuzziness, it is difficult to verify the accuracy of the quantitative weight coefficient of each element in the judgment matrix directly using the existing experience of the scorer, and the evaluation results lack rigorous persuasiveness.

[0009] In summary, the existing grid-connected wind farm operation performance evaluation methods have the following shortcomings:

[0010] At the evaluation index level, the existing grid-connected wind farm operation performance evaluation methods mainly focus on calculating the "grid side" indicators of wind turbines participating in the grid-connected system operation, and the evaluation of the operating characteristics of the wind turbines themselves on the "power supply side" is less considered. At the evaluation method level, the method of setting the weight coefficient of the influencing factor by artificial means has certain limitations, and there is no clear evaluation method for indicators that cannot be determined and have no quantitative description. Summary of the invention

[0011] Purpose of the invention: The purpose of the present invention is to provide a method and system for evaluating the operating performance of a grid-connected wind farm based on correlation analysis. By conducting data mining and comprehensive evaluation on the factors affecting the performance of wind turbines during operation, the quantitative relationship between performance evaluation indicators and various influencing factors is found, thereby overcoming the problem of less indicator coverage in existing evaluation standards.

[0012] Technical solution: A method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis of the present invention comprises:

[0013] Data acquisition: Establish a main monitoring area model in the wind farm operation performance evaluation system, and derive secondary models of wind farms, wind turbines, wind turbine impellers, transmission devices, and converters on this basis; obtain the information required for performance evaluation from the wind farm SCADA system, electric energy metering system, and meteorological station monitoring system to form historical sampling data based on time series;

[0014] Data preprocessing: filtering, smoothing and standardizing the historical sampling data based on time series to obtain the initial test data set for wind farm grid-connected performance evaluation;

[0015] Data correlation analysis: perform correlation analysis on the initial test data set, select the corresponding feature column data in the initial test data set according to the performance evaluation requirements to form a new test data set, and build the input feature vector of the wind farm operation performance evaluation based on this; mine the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, and select the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and build a regression model; perform regression analysis based on the regression model to obtain the quantitative description relationship between the feature vectors, and obtain the correction algorithm of the evaluation index or the new quantitative evaluation index; among which, the independent variable influencing factor is the factor that has an impact on the operation performance of the grid-connected wind farm, specifically including the wind farm ledger data, grid-connected operation data and wind energy resource data;

[0016] Performance evaluation: Substitute the actual operation data of the wind farm into the wind farm operation performance evaluation model composed of the new quantitative evaluation indicators and the revised quantitative evaluation indicators, calculate the values ​​of various performance indicators of the wind farm, and obtain the overall performance indicator evaluation results of the wind farm.

[0017] Furthermore, the initial test data set is subjected to correlation analysis, and the corresponding feature column data is selected from the initial test data set according to the performance evaluation requirements to form a new test data set, and the input feature vector of the wind farm operation performance evaluation is constructed based on this; the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set are mined, and the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable are selected according to the obtained correlation strength results to construct a regression model; regression analysis is performed based on the regression model to obtain the quantitative description relationship between the feature vectors, and the correction algorithm of the evaluation index or the new quantitative evaluation index is obtained, including:

[0018] Data normality check: Perform normality check on the test data set to check whether the sampling data in the test data set meets the normal distribution. After normality check, the test data set that meets the correlation analysis conditions is obtained;

[0019] Correlation coefficient calculation: Determine the calculation method of the correlation coefficient between the influencing factors in the test data set based on the normality test results; the correlation coefficient is used to determine the correlation strength between different feature columns in the test data set;

[0020] Remove abnormal result sets: Use the correlation coefficient to perform correlation analysis on all feature column data in the test data set. According to the correlation analysis results, select feature column data with correlation coefficients greater than a preset threshold in the test data set to form a new test data set, and construct the input feature vector for wind farm operation performance evaluation based on this.

[0021] Performance evaluation modeling: mining the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, selecting the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and constructing a regression model;

[0022] Solving regression equations: Considering the impact of wind turbine parameters, grid-connected operation data and wind energy environmental factors on wind turbine operating performance, multiple linear regression equations are established for each new test data set based on the regression model; by solving the multiple linear regression equations, the correlation between the influencing factors is converted into quantitative evaluation indicators for the operating performance of wind turbines and wind farms;

[0023] Overall verification: Conduct significance verification on the calculation method of correlation coefficient and the calculation results of multiple linear regression equation.

[0024] Furthermore, the test data set is subjected to normality check to check whether the sampling data in the test data set meets the normal distribution. After the normality check, the test data set that meets the correlation analysis conditions is obtained, including:

[0025] Calculate the percentiles of the sampled data in the test data set. The percentiles of the sampled data in the test data set and the percentiles of the normal distribution form a percentile comparison space. Preliminary judgment is made on whether the sampled data obeys the normal distribution according to the distribution of the sampled data points in the test data set.

[0026] Check the goodness of fit deviation between the sampling data in the test data set and the standard distribution, perform normality check on the SCADA historical sampling data in the test data set, and perform non-parametric test on the sampling data in the test data set using the KS hypothesis test method to check whether its assumption about the overall normal distribution is valid.

[0027] Furthermore, the goodness of fit deviation between the sampled data in the test data set and the standard distribution is checked, the normality check is performed on the SCADA historical sampled data in the test data set, and the KS hypothesis test method is used to perform a non-parametric test on the sampled data in the test data set to check whether the assumption about the overall normal distribution is valid, including:

[0028] After arranging the SCADA historical sampling data in the test data set in ascending order, the percentile marking points of the SCADA historical sampling data are obtained, and the mean and standard deviation of the percentile comparison space are calculated. Based on the degree of deviation between the measured point values ​​of the SCADA historical sampling data in the test data set and the standard deviation of the test data set, the distribution function of the test data set analysis sample is constructed:

[0029]

[0030] In the formula, σ is the standard deviation of historical sampling data; is the average value of the test data set.

[0031] Compare the deviations of the percentile marker interval points and their corresponding theoretical normal distribution cumulative function values, and compare the maximum absolute value with the KS critical value. When the maximum difference between the cumulative distribution function of the test data set and the cumulative function of the normal distribution is less than the critical value, the null hypothesis is considered to be established, otherwise the null hypothesis is rejected.

[0032] Calculate the maximum value of the difference between the cumulative distribution function of the test data set and the cumulative distribution function of the theoretical normal distribution, which is the KS statistic;

[0033] Set the significance level to determine whether the maximum value of the difference between the observed cumulative distribution function of the test data set and the cumulative distribution function of the theoretical normal distribution is a random error;

[0034] The calculated KS statistic is compared with the KS critical value. If the statistic KS is less than the KS critical value, the null hypothesis is considered to be established and the sampling data in the test data set basically obeys the normal distribution.

[0035] Furthermore, the calculation method of the correlation coefficient between the influencing factors in the test data set is determined according to the normality test results, including:

[0036] Normalization of test data set: A state data matrix is ​​constructed based on the test data set of potential influencing factors of the grid-connected wind farm operation performance evaluation indicators selected within a given time window. When there are large differences between the sampled data points in the state data matrix, the sampled data in the state data matrix are processed using the normalization method, specifically:

[0037]

[0038] In the formula, Represents the standardized value in the correlation analysis data matrix; x ij Represents the stored value of the jth potential influencing factor at the i-th sampling moment; μ j is the sampling average value of the jth potential influencing factor in a given time window; σ j For x ij The standard deviation of the jth potential influencing factor; the data in each column of the normalized state data matrix approximately obeys a normal distribution with a mean of 0 and a variance of 1;

[0039] When the selected test data set is close to normal distribution, the correlation analysis is performed on any two waveforms at a certain moment. The specific linear correlation degree is expressed by Pearson correlation coefficient, and its discretized expression is:

[0040]

[0041] In the formula, r is the correlation coefficient between the two variable sequences in a single analysis process; cov(X,Y), σ X , σ Y are the covariances of the two test data sets; x i ,y i are the values ​​of the i-th data point of a single analysis sampling data; are the average values ​​of two columns of sampled data respectively; n is the total amount of data contained in the sampled data window;

[0042] When the Pearson correlation coefficient of two sets of feature column data in the test data set is greater than the threshold value, it indicates that there is a significant correlation between the two sets of feature column data in the test data set.

[0043] Furthermore, the calculation method of the correlation coefficient between the influencing factors in the test data set is determined according to the normality test results, including:

[0044] If the data to be analyzed does not obey the normal distribution, it is necessary to use the rank score to realize the correlation check; when a data appears repeatedly, its rank is taken as the average of the two order ranks before and after, and the calculation method of the rank is restored after the data changes; the linear correlation degree of the two waveforms is expressed by the Spearman coefficient, and its discretization formula can be expressed as:

[0045]

[0046] Where, ρ is the Spearman correlation coefficient of the two sampling data sets; r i 、s i Represents the column rank of two columns of data; is the average of the two columns of data ranks; di represents the rank difference of the i-th pair of data; n is the total amount of sampled data for this analysis;

[0047] The larger the absolute value of the Spearman correlation coefficient, the higher the degree of correlation between the two types of data. The sign of the correlation coefficient determines the positive or negative correlation between the two types of data.

[0048] Furthermore, the calculation method of the correlation coefficient is checked for significance, including:

[0049] If the characteristic column of the test data set used to calculate the correlation coefficient follows a normal distribution and a new statistic t is constructed, then:

[0050]

[0051] In the formula, r represents the correlation coefficient of the calculated characteristic data column; n is the number of samples selected in the test data set; the statistic t obeys the t distribution with (n-2) degrees of freedom.

[0052] When the probability p value corresponding to the statistic t is less than the selected significance level, it is considered that there is a significant linear correlation between the feature columns of the selected test data set.

[0053] Furthermore, considering the influence of wind turbine parameters, wind turbine operating data and wind energy environment factors on the operating performance of wind turbines and wind farms, multiple linear regression equations are established for each of the new test data sets; by solving the multiple linear regression equations, the correlation between the influencing factors is converted into quantitative evaluation indicators for the operating performance of wind turbines and wind farms, including:

[0054] Establish a multiple linear regression equation:

[0055] Y=β 0 +β 1 X 1 +β 2 X 2 +…+β n X n +ε

[0056] In the formula, β 0 is a constant term; β i (i=1,2,3,……n) indicates that in other independent variables X i The average change in the dependent variable Y for each unit change of a specified independent variable X when it remains unchanged; ε represents the calculated residual;

[0057] In the process of solving the multivariate linear regression equation, its normalized equation system can be expressed as:

[0058] (X T X)B=X T Y

[0059] Where Y is the historical sampling data sequence of the dependent variable; B represents the historical sampling data sequence of the independent variable; and X is the coefficient matrix of the normalized equation group in the regression model.

[0060] Furthermore, the calculation results of the multivariate linear regression equation are tested for significance, including:

[0061] When all elements in the independent variable data sequence B are 0, the original linear correlation hypothesis is rejected, and the quantitative relationship is re-determined using the first-order linear or other methods; the calculation residuals are analyzed using the quantile diagram to check whether they approximately obey the normal distribution;

[0062] When the elements in the independent variable data sequence B are not all 0, the overall significance test of the multivariate linear regression equation is performed:

[0063]

[0064] In the formula, SS R Represents the sum of squares of the multiple linear regression equation; SS E Represents the residual sum of squares of the calculation results of the multivariate linear regression equation; Represents the i-th component of the predicted value of the regression equation; y i are the actual values ​​used to perform the regression analysis; represents the average of the actual values;

[0065] If the p-value of the multiple linear regression equation is less than the significance level, it means that the multiple linear regression equation is significant; perform a significance test on the regression coefficient. If the p-value is less than the significance level, the regression coefficient is significant; among them, there are:

[0066]

[0067] df R Represents the regression degrees of freedom, which is the number of independent variables in the multiple linear regression equation; df E represents the residual degrees of freedom and is determined by the following formula:

[0068] df E =df T -df R

[0069] df T =n-1

[0070] Among them, df T Represents the total degrees of freedom of the multiple linear regression equation; df R represents the regression degrees of freedom of the regression equation; n is the number of dimensions of the characteristic column vector.

[0071] The P value can be calculated by the following formula:

[0072] P=1-F X (x)

[0073] Among them, F X (x) is the cumulative distribution function value, which is characterized by F, df R 、df E Joint decision.

[0074] Based on the same inventive concept, a grid-connected wind farm operation performance evaluation system based on correlation analysis of the present invention comprises:

[0075] The data acquisition module is used to establish the main monitoring area model in the wind farm operation performance evaluation system, and derive the secondary models of the wind farm, wind turbine, wind turbine impeller, transmission device, and converter on this basis; obtain the information required for performance evaluation from the wind farm SCADA system, electric energy metering system, and meteorological station monitoring system to form historical sampling data based on time series;

[0076] The data preprocessing module is used to filter, smooth and standardize the historical sampling data based on the time series to obtain the initial test data set for the grid-connected performance evaluation of the wind farm;

[0077] The data correlation analysis module is used to perform correlation analysis on the initial test data set, select the corresponding feature column data in the initial test data set to form a new test data set according to the performance evaluation requirements, and construct the input feature vector of the wind farm operation performance evaluation based on this; mine the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, and select the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and construct a regression model; perform regression analysis based on the regression model to obtain the quantitative description relationship between the feature vectors, and obtain the correction algorithm of the evaluation index or the new quantitative evaluation index; wherein, the independent variable influencing factor is the factor that has an impact on the operation performance of the grid-connected wind farm, specifically including the wind farm ledger data, the grid-connected operation data and the wind energy resource data;

[0078] The performance evaluation module is used to substitute the actual operation data of the wind farm into the wind farm operation performance evaluation model composed of the new quantitative evaluation index and the revised quantitative evaluation index, calculate the values ​​of various performance indicators of the wind farm, and obtain the overall performance indicator evaluation result of the wind farm.

[0079] Beneficial effects: Compared with the prior art, the significant technical effects of the present invention are as follows:

[0080] Through correlation analysis of sampling data, we can discover potential individual factors that affect the evaluation indicators. By using multiple linear regression to solve the performance evaluation model, we can find the quantitative relationship between the influencing factors and the evaluation indicators, and provide technical support for the establishment of a wind farm operation performance evaluation index system.

[0081] Based on the actual needs of performance evaluation work, a performance evaluation model is constructed by screening factors with strong correlation and potential impact on the evaluation indicators according to the results of correlation analysis, providing a basis for studying unknown areas of performance evaluation that are difficult to explain and cannot be quantified for the time being.

[0082] For performance evaluation indicators with multiple calculation methods, the accuracy of their calculation methods can be verified through the calculation results of regression analysis. On this basis, the coefficients of influencing factors can be corrected to provide a reference for a more accurate description of the variable relationship.

[0083] The solved performance evaluation model and test data set can provide a quantitative basis for determining whether a wind turbine is in an abnormal operation or fault state, and provide a solution for determining and grading the health status of a wind turbine.

[0084] The overall framework of the performance index evaluation system has been improved. The evaluation scope is no longer limited to the classification method of indicators. As long as the physical quantity has an observed value, it can be used as a performance indicator for evaluation. The indicator calculation model with reasonable results can be included in the new indicator evaluation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 It is a flow chart of a method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis disclosed in an embodiment of the present invention;

[0086] Figure 2 It is a matrix diagram of a correlation analysis result of a performance indicator influencing factor disclosed in an embodiment of the present invention;

[0087] Figure 3 It is a structural schematic diagram of a grid-connected wind farm operation performance evaluation system based on correlation analysis disclosed in an embodiment of the present invention; DETAILED DESCRIPTION

[0088] The technical solution of the present invention is described in detail below in conjunction with specific implementation methods and the accompanying drawings of the specification.

[0089] Example 1

[0090] like Figure 1 As shown, a method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis of the present invention comprises the following steps:

[0091] S1. Data acquisition: Establish a main monitoring area model in the wind farm operation performance evaluation system, and derive secondary models such as wind farm, wind turbine, wind turbine impeller, transmission device, converter, etc. on this basis; obtain the information required for performance evaluation from data sources such as wind farm SCADA system, electric energy metering system, meteorological station monitoring system, etc., to form historical sampling data based on time series.

[0092] In this embodiment, data such as longitude, latitude, humidity, temperature, air pressure, wind speed, wind direction, and air density are obtained from the weather station monitoring system;

[0093] Obtain operating data such as voltage, current, active power, reactive power, grid-connected status (power generation, maintenance, power restriction, shutdown), real-time load, and planned power generation from the wind farm SCADA system; wind turbine operating parameter data such as rated capacity, model unit mark, cut-in / cut-out wind speed, etc.

[0094] Obtain pitch system data and converter data such as pitch angle, blade speed, cabinet temperature, and motor speed from the wind farm SCADA system;

[0095] Obtain transmission device data such as gearbox inlet oil temperature, outlet pressure, cabin temperature, cooling water temperature, etc. from the wind farm SCADA system;

[0096] Obtain electricity data such as online power consumption and field power consumption from the electric energy metering system.

[0097] In the wind farm SCADA system, the main monitoring area and weather station models are established to collect global data on the wind turbine side, wind farm grid side, wind energy resources, etc. On this basis, secondary models such as wind farm, unit transmission system, and electrical control system are established to improve the wind turbine account parameters and receive real-time uploaded data.

[0098] S2. Data preprocessing: Filter, smooth and standardize the historical sampling data based on time series to obtain the initial test data set for wind farm grid-connected performance evaluation.

[0099] In this embodiment, a first-order low-pass filter is used to filter the high-frequency random components collected and measured in the wind power SCADA system. These random components are usually generated by the fluctuations of the industrial field environment and the influence of the working conditions of the collection system. The filtering algorithm can be specifically expressed as:

[0100] X FIL (k+1)=X FIL (k)+(X RAW (k+1)-X FIL (k))*DT F (1)

[0101] Where k is the counter of the number of times the filtering algorithm is executed; X FIL Represents the value after filtering; X RAW Represents the raw data to be filtered; DT F Represents the discrete filter factor, which ranges from 0 to 1.

[0102] The filtering algorithm is based on SCADA data and can reduce the interference of invalid data when combined with the measurement point and measurement validity verification. In general, it can be judged according to the following attributes: a) SCADA measurement has a bad quality mark; b) The measurement exceeds the normal specified range; c) The measurement does not change within the specified time; d) There is a sudden change in the measurement data within two consecutive indicator evaluation cycles.

[0103] Perform preprocessing operations such as filtering and smoothing on the historical sampling data based on time series to reduce the impact of noise and ensure the accuracy and comparability of the data.

[0104] Taking the operating conditions of wind turbines as reference, several types of data grouping are established, including stable operation of wind turbines, operating condition switching of wind turbines and faulty operation of wind turbines, and performance evaluation test data sets are obtained as sample data for subsequent analysis.

[0105] S3. Data correlation analysis: perform correlation analysis on the initial test data set, select corresponding feature column data in the initial test data set according to the performance evaluation requirements to form a new test data set, and construct the input feature vector of wind farm operation performance evaluation based on this; mine the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vector of each wind farm operation performance evaluation in the new test data set, and select the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and construct a regression model; perform regression analysis based on the regression model to obtain the quantitative description relationship between the feature vectors, and obtain the correction algorithm of the evaluation index or the new quantitative evaluation index; among which, the independent variable influencing factor is the factor that has an impact on the operation performance of the grid-connected wind farm, specifically including wind farm ledger data, grid-connected operation data and wind energy resource data.

[0106] Among them, the preset grid-connected wind farm operation performance evaluation indicators include basic indicators, resource indicators, electricity indicators, energy consumption indicators, operation indicators, maintenance indicators and energy storage indicators.

[0107] Among them, the influencing factor is the independent variable x, the evaluation index is the dependent variable y, and the result of the correlation analysis is to combine the factors with greater correlation strength into a function expression y=f(x) of the independent variable and the dependent variable.

[0108] The specific process of step S3 is as follows:

[0109] S3.1. Data normality check: Perform normality check on the test data set to check whether the sampling data in the test data set meets the normal distribution. After normality check, the test data set that meets the correlation analysis conditions is obtained. Specifically, the following steps are included:

[0110] S3.1.1. Calculate the percentiles of the sampling data in the test data set. The percentiles of the sampling data in the test data set and the percentiles of the normal distribution form a percentile comparison space. Based on the distribution of the sampling data points in the test data set, preliminarily determine whether the sampling data obeys the normal distribution.

[0111] When the sample size of SCADA historical data in the test data set is small or the wind farm is in steady-state operation, the following three methods can be used to perform normality verification:

[0112] 1) Use the quartile and standard deviation to make a rough first-round judgment. When the ratio of the two is greater than a given threshold, it is approximately considered that the test data set satisfies the normal distribution.

[0113] 2) Calculate the cumulative probability and percentile of the test data, and use a histogram to view the comparison between the calculated value and the theoretical value of the standard distribution.

[0114] 3) Calculate the kurtosis and skewness values ​​of the test data, and conduct a combined quantitative investigation of the symmetry and deviation of the test data:

[0115]

[0116] In the formula, Z-Score skew 、Z-Score curtosis Respectively represent the Z score index of the test data skewness and kurtosis; V skew 、V curtosis Respectively represent the skewness and kurtosis values ​​of the test data; σ skew , σ curtosis They represent the standard deviation of the skewness and kurtosis values ​​of this group of data respectively.

[0117] S3.1.2. Check the goodness of fit deviation between the sampling data in the test data set and the standard distribution, perform normality check on the SCADA historical sampling data in the test data set, and perform non-parametric test on the sampling data in the test data set using the KS hypothesis test method to check whether its assumption about the overall normal distribution is valid.

[0118] Since the sampling data involves a relatively long period of time, a certain period of time is selected for analysis.

[0119] The specific process of step S3.1.2 is as follows:

[0120] S3.1.2.1. Arrange the SCADA historical sampling data in the test data set in ascending order, find the percentile mark points of the SCADA historical sampling data, calculate the mean and standard deviation of the percentile comparison space, and construct the distribution function of the test data set analysis sample based on the degree of deviation between the SCADA historical sampling data point value in the test data set and the standard deviation of the test data set:

[0121]

[0122] In the formula, σ is the standard deviation of historical sampling data; is the average value of the test data set

[0123] S3.1.2.2. Compare the deviations of the percentile marker interval points and their corresponding theoretical normal distribution cumulative function values, and compare the maximum absolute value with the KS critical value. When the maximum difference between the cumulative distribution function of the test data set and the cumulative function of the normal distribution is less than the critical value, the null hypothesis is considered to be established, otherwise the null hypothesis is rejected;

[0124] Among them, the corresponding theoretical normal distribution cumulative distribution function can be obtained according to the mean value and standard deviation of the characteristic data column.

[0125] It is worth noting that the absolute value here refers to the absolute value of the deviation between each data in the feature data column and the corresponding theoretical normal distribution cumulative function value. The number of deviation values ​​and the absolute value of the deviation data will be the same as the number of data in the feature data column.

[0126] S3.1.2.3. Calculate the maximum value of the difference between the cumulative distribution function of the test data set and the cumulative distribution function of the theoretical normal distribution, which is the KS statistic;

[0127] Each time a performance evaluation analysis is conducted, a batch of data is taken from the historical sampling data to form a test data set, and from a longitudinal perspective, multiple test data sets will be formed.

[0128] S3.1.2.4. Set a significance level to determine whether the maximum value of the difference between the observed cumulative distribution function of the test data set and the cumulative distribution function of the theoretical normal distribution is a random error;

[0129] S3.1.2.5. Compare the calculated KS statistic with the KS critical value. If the statistic KS is less than the KS critical value, the null hypothesis is considered to be true and the sampled data in the test data set basically obey the normal distribution.

[0130] To ensure data accuracy, it is necessary to compare and verify the sampling properties in the test data set: 1) The sampling time intervals of each evaluation quantity must be kept consistent. For individual missing data, they can be obtained through interpolation algorithms or moving time windows to ensure that each data point has a sampling record; 2) Check the difference between the sampling value and the mean in each test data set. If the difference is too large, it is necessary to determine whether there are wind turbines in a working condition switching state or a fault state.

[0131] The KS hypothesis is the Kolmogorov-Smirnov test, which is a hypothesis testing method used to check whether the distribution form of sample data obeys the normal distribution.

[0132] S3.2, Correlation coefficient calculation: Determine the calculation method of the correlation coefficient between the influencing factors in the test data set based on the normality test results; the correlation coefficient is used to determine the correlation strength between different feature columns in the test data set. The details are as follows:

[0133] Normalization of test data set: A state data matrix is ​​constructed based on the test data set of potential influencing factors of the grid-connected wind farm operation performance evaluation indicators selected within a given time window. When there are large differences between the sampled data points in the state data matrix, the sampled data in the state data matrix are processed using the normalization method, specifically:

[0134]

[0135] In the formula, Represents the standardized value in the correlation analysis data matrix; x ij Represents the stored value of the jth potential influencing factor at the i-th sampling moment; μ j is the sampling average value of the jth potential influencing factor in a given time window; σ j For x ij The standard deviation of the jth potential influencing factor; the data in each column of the normalized state data matrix approximately obeys a normal distribution with a mean of 0 and a variance of 1;

[0136] When the selected test data set is close to normal distribution, the correlation analysis is performed on any two waveforms at a certain moment. The specific linear correlation degree is expressed by Pearson correlation coefficient, and its discretized expression is:

[0137]

[0138] In the formula, r is the correlation coefficient between the two variable sequences in a single analysis process; cov(X,Y), σ X , σ Y are the covariances of the two test data sets; x i ,y iare the values ​​of the i-th data point of a single analysis sampling data; are the average values ​​of two columns of sampled data respectively; n is the total amount of data contained in the sampled data window;

[0139] When the Pearson correlation coefficient of two sets of feature column data in the test data set is greater than the threshold value, it indicates that there is a significant correlation between the two sets of feature column data in the test data set.

[0140] According to the correlation calculation results between any two modules in the test data set, the strength of the correlation between the two is determined. Generally, when the Pearson correlation coefficient |r|≤0.3, it means that the correlation between the two sets of data is very low; when 0.3<|r|≤0.5, it means that the correlation between the feature columns of the two test data sets is low; when 0.5<|r|≤0.8, it means that the correlation between the feature columns of the two test data sets is high; when 0.8<|r|≤1, it means that the correlation between the feature columns of the two test data sets is very high. The value range of r is [-1,1]. When r=1, it means that the two sets of test data are completely positively correlated; when r=-1, it means that the two sets of test data are completely negatively correlated.

[0141] If the data to be analyzed does not follow a normal distribution, it is necessary to use the rank score to implement correlation verification. When a data appears multiple times, its rank is taken as the average of the two order ranks before and after, and the calculation method of the rank is restored after the data changes. The linear correlation degree of the two waveforms is expressed by the Spearman correlation coefficient, and its discretization formula can be expressed as:

[0142]

[0143]

[0144] Where, ρ is the Spearman correlation coefficient of the two sampling data sets; r i 、s i Represents the column rank of two columns of data; is the average of the two columns of data ranks; d i represents the rank difference of the i-th pair of data; n is the total amount of sampled data for this analysis.

[0145] If the Spearman correlation coefficient is selected to evaluate the correlation between the influencing factors of the performance index, the larger the absolute value of the Spearman correlation coefficient, the higher the correlation between the characteristic columns of the selected test data set. The positive or negative sign of the correlation coefficient determines the positive or negative correlation of the analyzed data.

[0146] The above correlation coefficient is calculated based on the sample data within a given time window, and the resulting correlation coefficient is a sample correlation coefficient. Due to the randomness of sampling and the limitation of sample quantity, the sample correlation coefficient cannot be directly used to indicate whether there is a significant linear correlation between two feature columns.

[0147] By mining and calculating the correlation between the grid-connected wind farm operating performance indicators and potential influencing factors, the variable factors that significantly affect the performance indicators are found.

[0148] Compare the deviations of the normality check results with the given threshold and determine the calculation method of the correlation coefficients of the influencing factors to be analyzed in the experimental data set.

[0149] Determine the data columns where the performance evaluation indicators of wind turbines or wind farms and potential influencing factors are located, calculate the corresponding correlation coefficients, and determine whether there is a strong correlation between the selected sampling data columns as a basis for further analysis and judgment.

[0150] S3.3. Remove abnormal result sets: Use the correlation coefficient to perform correlation analysis on all feature column data in the test data set, and select feature column data with correlation coefficients greater than a preset threshold in the test data set according to the correlation analysis results to form a new test data set.

[0151] By eliminating abnormal data sets or data sets that cannot be explained temporarily in the correlation analysis results, we can obtain a test data set that approximately satisfies the linear correlation, and on this basis, we can construct the input feature vector for wind farm operation performance evaluation.

[0152] S3.4, Performance evaluation modeling: Mining the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, and selecting the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and constructing a regression model.

[0153] The regression model comprehensively considers the impact of factors such as unit parameters, operating data and wind energy environment on the operating performance of units and wind farms.

[0154] S3.5, solve the regression equation: Consider the influence of wind turbine parameters, grid-connected operation data and wind energy environmental factors on the operation performance of wind turbines, and establish multiple linear regression equations for new test data sets one by one based on the regression model; by solving the multiple linear regression equation, the correlation between the influencing factors is converted into quantitative evaluation indicators for the operation performance of wind turbines and wind farms. The details are as follows:

[0155] Multiple linear regression analysis is used to determine the quantitative relationship between the various influencing factors of wind turbine performance evaluation indicators. Multiple linear regression equation is established:

[0156] Y=β 0 +β 1 X 1 +β 2 X 2 +…+β n X n +ε

[0157] In the formula, β 0 is a constant term; β i (i=1,2,3,……n) indicates that in other independent variables X i The average change in the dependent variable Y for each unit change in a specified independent variable X when it remains unchanged; ε represents the calculated residual, which is the difference between the true value and the estimated value of the dependent variable and is a random variable.

[0158] In the process of solving the multivariate linear regression equation, its normalized equation system can be expressed as:

[0159] (X T X)B=X T Y

[0160] Where Y is the historical sampling data sequence of the dependent variable; B represents the historical sampling data sequence of the independent variable; and X is the coefficient matrix of the normalized equation group in the regression model.

[0161] The data sets that are determined to be nonlinear relationships based on actual on-site operation experience of wind farms are converted into linear relationships through variable transformation and re-substituted into step S3.5 to calculate the quantitative relationship between the research data sets.

[0162] Calculate the exact quantitative relationship between relevant wind farm performance indicators and significant influencing factors, and perform an overall check on the calculation results.

[0163] S3.6. Overall verification: Conduct significance verification on the calculation method of the correlation coefficient and the calculation results of the multivariate linear regression equation.

[0164] The calculation method of the correlation coefficient is verified for significance, including:

[0165] After the above process, if the characteristic column of the test data set used to calculate the correlation coefficient obeys the normal distribution, a new statistic t is constructed, then:

[0166]

[0167] In the formula, r represents the correlation coefficient of the calculated characteristic data column; n is the number of samples selected in the test data set. The statistic t follows a t distribution with (n-2) degrees of freedom.

[0168] When the probability p value corresponding to the statistic t is less than the selected significance level, it is considered that there is a significant linear correlation between the feature columns of the selected test data set.

[0169] The calculation results of the multivariate linear regression equation are tested for significance, and the variables in the impact factor data set that are related to the performance evaluation indicators are determined, including:

[0170] (1) When all elements in the independent variable data sequence B are 0, the original linear correlation hypothesis is rejected and the quantitative relationship is re-determined using first-order linear or other methods; the calculated residuals are analyzed using quantile plots to check whether they approximately obey the normal distribution;

[0171] By analyzing the calculated residuals of the regression model, we can test whether the basic assumptions of linear regression are met. If the normal probability plot of the residuals does not have a large drift, it means that the previous assumptions are not obviously wrong.

[0172] (2) When the elements in the independent variable data sequence B are not all zero, perform an overall significance check on the multivariate linear regression equation:

[0173]

[0174] In the formula, SS R Represents the sum of squares of the multiple linear regression equation; SS E Represents the residual sum of squares of the calculation results of the multivariate linear regression equation; Represents the i-th component of the predicted value of the regression equation; y i represents the actual value used to perform the regression analysis; Indicates the average of the actual values.

[0175] (3) If the p-value of the multiple linear regression equation is less than the significance level, it means that the multiple linear regression equation is significant; perform a significance test on the regression coefficient. If the p-value is less than the significance level, the regression coefficient is significant; among them, there are:

[0176]

[0177] Where, df R Represents the regression degrees of freedom, which is the number of independent variables in the multiple linear regression equation; df E represents the residual degrees of freedom and is determined by the following formula:

[0178] df E =df T -df R

[0179] df T =n-1

[0180] Among them, dfT Represents the total degrees of freedom of the multiple linear regression equation; df R represents the regression degrees of freedom of the regression equation; n is the number of dimensions of the characteristic column vector.

[0181] The P value can be calculated by the following formula:

[0182] P=1-F X (x)

[0183] Among them, F X (x) is the cumulative distribution function value, which is characterized by F, df R 、df E Joint decision.

[0184] S4. Performance evaluation: Substitute the actual operation data of the wind farm into the wind farm operation performance evaluation model composed of the new quantitative evaluation indicators and the revised quantitative evaluation indicators, calculate the values ​​of various performance indicators of the wind farm, and obtain the overall performance indicator evaluation results of the wind farm.

[0185] By solving the multivariate linear regression, we can obtain the specific functional relationship between the evaluation indicators and the influencing factors. The collection of these relationship is the wind farm operation performance evaluation model.

[0186] Performance evaluation includes operational analysis, economic operation and statistical analysis. The statistical analysis submodule includes statistical indicator analysis, key indicator query, focus information benchmarking and other functions.

[0187] Operation analysis submodule, including equipment efficiency, operation and maintenance rates, and wind energy resources;

[0188] The economic operation submodule includes theoretical power balance analysis, power loss analysis, economic operation overview, station operation comparison and other functions.

[0189] The method for evaluating the health status of a wind turbine during grid-connected operation includes: when a wind turbine fails or operates abnormally, the correlation between its parameters will also change. The correlation coefficients between the parameters during normal operation can be compared with the correlation coefficients during abnormal conditions. The fault interval that causes the wind turbine to be in an unhealthy state can be located based on the fault characteristic parameters with large deviations.

[0190] The wind farm operation indicators are extracted and calculated by category to obtain the data source of wind farm grid-connected performance evaluation indicators.

[0191] The data relationships with quantitative descriptions will be found and added to the relevant categories to improve the existing wind turbine operation performance evaluation index system.

[0192] Improve the grid-connected wind farm operation performance index evaluation system. According to the specific quantitative representation method of wind farm operation performance evaluation index, the actual operation data of the subsequent connected wind farm will be entered into the evaluation model to calculate and evaluate the various performance indicators of the wind farm operation, and the evaluation results of each evaluation index will be obtained to provide data support for the economical and safe operation and daily maintenance of the wind farm.

[0193] The present invention comprehensively analyzes the coupling between wind turbine operation data and environmental resources and unit parameters, uses correlation coefficient method and regression analysis method to conduct data mining and comprehensive evaluation on the factors affecting the performance of wind turbines during operation, and finds the quantitative relationship between performance indicators and various factors from the evaluation method level. This method overcomes the problem of less coverage of indicators in existing evaluation standards, and can establish a more comprehensive and in-depth evaluation system for wind turbine operation performance indicators; it provides a scientific basis for the efficient management of wind farm operation and maintenance, fault identification and optimization of wind turbines, and also provides technical support for the subsequent improvement of relevant industry standards.

[0194] In addition, the method proposed in the present invention can solve some of the shortcomings of the document 1 from a mathematical perspective: the correlation analysis method is a mathematical method, which has no special requirements on the physical meaning of the characteristic vector. When the measurement conditions permit, the characteristic column evaluation index can select the "grid side" index of the wind turbine group participating in the system operation, or select the operating parameters and indicators of the wind turbine group based on its own operating characteristics for analysis; it can be an evaluation quantity of an existing calculation formula, or an evaluation quantity that is not currently quantitatively described but actively explored and analyzed. The results of the analysis can provide certain data support for the improvement of the indicator system.

[0195] Example 2

[0196] like Figure 3 As shown, a grid-connected wind farm operation performance evaluation system based on correlation analysis of the present invention comprises:

[0197] The data acquisition module is used to establish the main monitoring area model in the wind farm operation performance evaluation system, and derive the secondary models of the wind farm, wind turbine, wind turbine impeller, transmission device, and converter on this basis; obtain the information required for performance evaluation from the wind farm SCADA system, electric energy metering system, and meteorological station monitoring system to form historical sampling data based on time series;

[0198] The data preprocessing module is used to filter, smooth and standardize the historical sampling data based on the time series to obtain the initial test data set for the grid-connected performance evaluation of the wind farm;

[0199] The data correlation analysis module is used to perform correlation analysis on the initial test data set, select the corresponding feature column data in the initial test data set to form a new test data set according to the performance evaluation requirements, and construct the input feature vector of the wind farm operation performance evaluation based on this; mine the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, and select the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and construct a regression model; perform regression analysis based on the regression model to obtain the quantitative description relationship between the feature vectors, and obtain the correction algorithm of the evaluation index or the new quantitative evaluation index; wherein, the independent variable influencing factor is the factor that has an impact on the operation performance of the grid-connected wind farm, specifically including the wind farm ledger data, the grid-connected operation data and the wind energy resource data;

[0200] The performance evaluation module is used to substitute the actual operation data of the wind farm into the wind farm operation performance evaluation model composed of the new quantitative evaluation index and the revised quantitative evaluation index, calculate the values ​​of various performance indicators of the wind farm, and obtain the overall performance indicator evaluation result of the wind farm.

[0201] In an optional embodiment, a method for evaluating the operating performance of a grid-connected wind farm based on correlation analysis includes: a) establishing a main monitoring area model, and deriving secondary models of the wind farm, wind turbine, wind turbine impeller, transmission device, and converter on this basis; obtaining information required for performance evaluation to form historical sampling data based on time series; b) filtering, smoothing, and standardizing the historical sampling data based on time series to obtain an initial test data set for wind farm grid-connected performance evaluation; c) performing correlation analysis on the initial test data set, using correlation coefficient method and regression analysis method to perform data mining and comprehensive evaluation on factors affecting the performance of the wind turbine during operation, and finding the quantitative relationship between performance indicators and various factors from the evaluation method level; d) substituting the actual operating data of the wind farm into the wind farm operating performance evaluation model, and calculating the values ​​of various performance indicators of the wind farm according to the quantitative expression of indicator evaluation obtained in the above process to obtain the overall performance indicator evaluation result of the wind farm.

Claims

1. A method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis, characterized in that: include: Data acquisition: Establish a main monitoring area model in the wind farm operation performance evaluation system, and derive secondary models of wind farms, wind turbines, wind turbine impellers, transmission devices, and converters on this basis; obtain the information required for performance evaluation from the wind farm SCADA system, electric energy metering system, and meteorological station monitoring system to form historical sampling data based on time series; Data preprocessing: filtering, smoothing and standardizing the historical sampling data based on time series to obtain the initial test data set for wind farm grid-connected performance evaluation; Data correlation analysis: perform correlation analysis on the initial test data set, select the corresponding feature column data in the initial test data set according to the performance evaluation requirements to form a new test data set, and build the input feature vector of the wind farm operation performance evaluation based on this; mine the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, and select the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and build a regression model; perform regression analysis based on the regression model to obtain the quantitative description relationship between the feature vectors, and obtain the correction algorithm of the evaluation index or the new quantitative evaluation index; among which, the independent variable influencing factor is the factor that has an impact on the operation performance of the grid-connected wind farm, specifically including the wind farm ledger data, grid-connected operation data and wind energy resource data; Performance evaluation: Substitute the actual operation data of the wind farm into the wind farm operation performance evaluation model composed of the new quantitative evaluation indicators and the revised quantitative evaluation indicators, calculate the values ​​of various performance indicators of the wind farm, and obtain the overall performance indicator evaluation results of the wind farm.

2. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 1, characterized in that: Conduct correlation analysis on the initial test data set, select corresponding feature column data in the initial test data set to form a new test data set according to the performance evaluation requirements, and build a wind farm operation performance evaluation input feature vector based on this; mine the correlation between the performance evaluation index to be analyzed and the new test data set, and the correlation between the input feature vectors of the wind farm operation performance evaluation in the new test data set, and select the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and build a regression model; Based on the regression model, regression analysis is performed to obtain the quantitative description relationship between the feature vectors, and the correction algorithm of the evaluation index or the new quantitative evaluation index is obtained, including: Data normality check: Perform normality check on the test data set to check whether the sampling data in the test data set meets the normal distribution. After normality check, the test data set that meets the correlation analysis conditions is obtained; Correlation coefficient calculation: Determine the calculation method of the correlation coefficient between the influencing factors in the test data set based on the normality test results; the correlation coefficient is used to determine the correlation strength between different feature columns in the test data set; Remove abnormal result sets: Use the correlation coefficient to perform correlation analysis on all feature column data in the test data set. According to the correlation analysis results, select feature column data with correlation coefficients greater than a preset threshold in the test data set to form a new test data set, and construct the input feature vector for wind farm operation performance evaluation based on this. Performance evaluation modeling: mining the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, selecting the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and constructing a regression model; Solving the regression equation: Considering the influence of wind turbine parameters, grid-connected operation data and wind energy environment factors on the operation performance of wind turbines, based on the regression model obtained in the previous step, multiple linear regression equations are established for the new test data sets one by one; by solving the multiple linear regression equation, the correlation between the influencing factors is converted into a quantitative evaluation index for the operation performance of wind turbines and wind farms; Overall verification: Conduct significance verification on the calculation method of correlation coefficient and the calculation results of multiple linear regression equation.

3. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 2 is characterized in that: Perform normality check on the test data set to check whether the sampling data in the test data set meets the normal distribution. After normality check, the test data set that meets the correlation analysis conditions is obtained, including: Calculate the percentiles of the sampled data in the test data set. The percentiles of the sampled data in the test data set and the percentiles of the normal distribution form a percentile comparison space. Preliminary judgment is made on whether the sampled data obeys the normal distribution according to the distribution of the sampled data points in the test data set. Check the goodness of fit deviation between the sampling data in the test data set and the standard distribution, perform normality check on the SCADA historical sampling data in the test data set, and perform non-parametric test on the sampling data in the test data set using the KS hypothesis test method to check whether its assumption about the overall normal distribution is valid.

4. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 3 is characterized in that: Check the goodness of fit deviation between the sampled data in the test data set and the standard distribution, perform normality check on the SCADA historical sampled data in the test data set, and perform non-parametric test on the sampled data in the test data set by KS hypothesis test method to check whether the assumption about the overall normal distribution is valid, including: After arranging the SCADA historical sampling data in the test data set in ascending order, the percentile marking points of the SCADA historical sampling data are obtained, and the mean and standard deviation of the percentile comparison space are calculated. Based on the degree of deviation between the measured point values ​​of the SCADA historical sampling data in the test data set and the standard deviation of the test data set, the distribution function of the test data set analysis sample is constructed: In the formula, σ is the standard deviation of historical sampling data; is the average value of the test data set. Compare the deviations of the percentile marker interval points and their corresponding theoretical normal distribution cumulative function values, and compare the maximum absolute value with the KS critical value. When the maximum difference between the cumulative distribution function of the test data set and the cumulative function of the normal distribution is less than the critical value, the null hypothesis is considered to be established, otherwise the null hypothesis is rejected. Calculate the maximum value of the difference between the cumulative distribution function of the test data set and the cumulative distribution function of the theoretical normal distribution, which is the KS statistic; Set the significance level to determine whether the maximum value of the difference between the observed cumulative distribution function of the test data set and the cumulative distribution function of the theoretical normal distribution is a random error; The calculated KS statistic is compared with the KS critical value. If the statistic KS is less than the KS critical value, the null hypothesis is considered to be established and the sampling data in the test data set basically obeys the normal distribution.

5. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 2, characterized in that: The calculation method of the correlation coefficient between the influencing factors in the test data set is determined according to the normality test results, including: Normalization of test data set: A state data matrix is ​​constructed based on the test data set of potential influencing factors of the grid-connected wind farm operation performance evaluation indicators selected within a given time window. When there are large differences between the sampled data points in the state data matrix, the sampled data in the state data matrix are processed using the normalization method, specifically: In the formula, Represents the standardized value in the correlation analysis data matrix; x ij Represents the stored value of the jth potential influencing factor at the i-th sampling moment; μ j is the sampling average value of the jth potential influencing factor in a given time window; σ j For x ij The standard deviation of the jth potential influencing factor; the data in each column of the normalized state data matrix approximately obeys a normal distribution with a mean of 0 and a variance of 1; When the selected test data set is close to normal distribution, the correlation analysis is performed on any two waveforms at a certain moment. The specific linear correlation degree is expressed by Pearson correlation coefficient, and its discretized expression is: In the formula, r is the correlation coefficient between the two variable sequences in a single analysis process; cov(X,Y), σ X , σ Y are the covariances of the two test data sets; x i ,y i are the values ​​of the i-th data point of a single analysis sampling data; are the average values ​​of two columns of sampled data respectively; n is the total amount of data contained in the sampled data window; When the Pearson correlation coefficient of two sets of feature column data in the test data set is greater than the threshold value, it indicates that there is a significant correlation between the two sets of feature column data in the test data set.

6. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 2, characterized in that: The calculation method of the correlation coefficient between the influencing factors in the test data set is determined according to the normality test results, including: If the data to be analyzed does not obey the normal distribution, it is necessary to use the rank score to realize the correlation check; when a data appears repeatedly, its rank is taken as the average of the two order ranks before and after, and the calculation method of the rank is restored after the data changes; the linear correlation degree of the two waveforms is expressed by the Spearman coefficient, and its discretization formula can be expressed as: Where, ρ is the Spearman correlation coefficient of the two sampling data sets; r i 、s i Represents the column rank of two columns of data; is the average of the two columns of data ranks; d i represents the rank difference of the i-th pair of data; n is the total amount of sampled data for this analysis; The larger the absolute value of the Spearman correlation coefficient, the higher the degree of correlation between the two types of data. The sign of the correlation coefficient determines the positive or negative correlation between the two types of data.

7. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 2, characterized in that: The calculation method of the correlation coefficient is verified for significance, including: If the characteristic column of the test data set used to calculate the correlation coefficient follows a normal distribution and a new statistic t is constructed, then: In the formula, r represents the correlation coefficient of the calculated characteristic data column; n is the number of samples selected in the test data set; the statistic t obeys the t distribution with (n-2) degrees of freedom. When the probability p value corresponding to the statistic t is less than the selected significance level, it is considered that there is a significant linear correlation between the feature columns of the selected test data set.

8. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 2, characterized in that: Considering the influence of wind turbine parameters, grid-connected operation data and wind energy environment factors on the operation performance of wind turbines, multiple linear regression equations are established for each new test data set. By solving the multiple linear regression equations, the correlation between the influencing factors is converted into quantitative evaluation indicators for the operation performance of wind turbines and wind farms, including: Establish a multiple linear regression equation: Y=β0+β1X1+β2X2+…+β n X n +e Where β0 is a constant term; β i (i=1,2,3,……n) indicates that in other independent variables X i The average change in the dependent variable Y for each unit change of a specified independent variable X when it remains unchanged; ε represents the calculated residual; In the process of solving the multivariate linear regression equation, its normalized equation system can be expressed as: (X T X)B=X T Y Where Y is the historical sampling data sequence of the dependent variable; B represents the historical sampling data sequence of the independent variable; and X is the coefficient matrix of the normalized equation group in the regression model.

9. The method for evaluating the operation performance of a grid-connected wind farm based on correlation analysis according to claim 2, characterized in that: The calculation results of the multivariate linear regression equation are tested for significance, including: When all elements in the independent variable data sequence B are 0, the original linear correlation hypothesis is rejected, and the quantitative relationship is re-determined using the first-order linear or other methods; the calculation residuals are analyzed using the quantile diagram to check whether they approximately obey the normal distribution; When the elements in the independent variable data sequence B are not all 0, the overall significance test of the multivariate linear regression equation is performed: In the formula, SS R Represents the sum of squares of the multiple linear regression equation; SS E Represents the residual sum of squares of the calculation results of the multivariate linear regression equation; Represents the i-th component of the predicted value of the regression equation; y i represents the actual value used to perform the regression analysis; Indicates the average of the actual values. If the p-value of the multiple linear regression equation is less than the significance level, it means that the multiple linear regression equation is significant; perform a significance test on the regression coefficient. If the p-value is less than the significance level, the regression coefficient is significant; among them, there are: df R Represents the regression degrees of freedom, which is the number of independent variables in the multiple linear regression equation; df E represents the residual degrees of freedom and is determined by the following formula: df E =df T -df R df T =n-1 Among them, df T Represents the total degrees of freedom of the multiple linear regression equation; df R represents the regression degrees of freedom of the regression equation; n is the number of dimensions of the characteristic column vector. The P value can be calculated by the following formula: P=1-F X (x) Among them, F X (x) is the cumulative distribution function value, which is characterized by F, df R 、df E Joint decision.

10. A grid-connected wind farm operation performance evaluation system based on correlation analysis, characterized in that: include: The data acquisition module is used to establish the main monitoring area model in the wind farm operation performance evaluation system, and derive the secondary models of the wind farm, wind turbine, wind turbine impeller, transmission device, and converter on this basis; obtain the information required for performance evaluation from the wind farm SCADA system, electric energy metering system, and meteorological station monitoring system to form historical sampling data based on time series; The data preprocessing module is used to filter, smooth and standardize the historical sampling data based on the time series to obtain the initial test data set for the grid-connected performance evaluation of the wind farm; The data correlation analysis module is used to perform correlation analysis on the initial test data set, select the corresponding feature column data in the initial test data set to form a new test data set according to the performance evaluation requirements, and construct the input feature vector of the wind farm operation performance evaluation based on this; mine the correlation between the performance evaluation index to be analyzed and the new test data set, and the input feature vectors of the wind farm operation performance evaluation in the new test data set, and select the influencing factor sequence as the independent variable and the performance evaluation index as the dependent variable according to the obtained correlation strength results, and construct a regression model; perform regression analysis based on the regression model to obtain the quantitative description relationship between the feature vectors, and obtain the correction algorithm of the evaluation index or the new quantitative evaluation index; wherein, the independent variable influencing factor is the factor that has an impact on the operation performance of the grid-connected wind farm, specifically including the wind farm ledger data, the grid-connected operation data and the wind energy resource data; The performance evaluation module is used to substitute the actual operation data of the wind farm into the wind farm operation performance evaluation model composed of the new quantitative evaluation index and the revised quantitative evaluation index, calculate the values ​​of various performance indicators of the wind farm, and obtain the overall performance indicator evaluation result of the wind farm.

Images