Method and system for analyzing influence of high proportion of distributed photovoltaic access to power distribution network

Abstract

This invention relates to a method and system for analyzing the impact of high-proportion distributed photovoltaic (PV) grid integration into a distribution network. The method includes the following steps: constructing a high-proportion distributed PV probability density function based on the basic data of the high-proportion distributed PV to be integrated; calculating the Spearman correlation coefficients between the power variables of the high-proportion distributed PV to construct a first Spearman correlation matrix; obtaining a high-proportion distributed PV power sampling matrix using a Latin hypercube sampling method based on a genetic algorithm, wherein the error between the second Spearman correlation matrix of this matrix and the first Spearman correlation matrix is ​​less than a set threshold; acquiring distribution network data; calculating the electrical parameters after the high-proportion distributed PV is integrated into the distribution network based on the sampling matrix; and obtaining the impact analysis results of the high-proportion distributed PV integration into the distribution network based on the electrical parameters. Compared with existing technologies, this invention has advantages such as improving the reliability of distribution network operation.

Description

Technical Field

[0001] This invention relates to the field of distribution network operation optimization, and in particular to a method and system for analyzing the impact of high-proportion distributed photovoltaic (PV) grid integration on distribution networks. Background Technology

[0002] With increasing public awareness of environmental issues, solar energy, due to its advantages such as low pollution, abundant resources, and renewability, has become a powerful way to promote energy transition and protect our planet. Distributed photovoltaic (PV) power generation, a type of PV power generation, offers advantages such as significantly reducing energy losses caused by long-distance transmission, resulting in efficient energy use and high flexibility. As the demand for clean energy gradually increases, the application of distributed PV is becoming increasingly widespread.

[0003] Meanwhile, the output of distributed photovoltaic (PV) power is easily affected by weather factors such as solar irradiance, resulting in varying power outputs at different times of the year and exhibiting randomness and volatility. Therefore, the integration of distributed PV can hinder the safe operation of the distribution network to some extent, posing challenges to its normal functioning. When a high proportion of distributed PV is incorporated, the power quality of the distribution network decreases, its network structure undergoes significant changes, many electrical indicators change, and its power supply reliability is also affected. Traditional distribution network planning and operation methods are no longer applicable under high-proportion distributed PV integration. How to ensure the normal and reliable operation of the distribution network under high-proportion distributed PV integration is a pressing technical problem that requires solutions and has significant theoretical research value and engineering practical significance. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a method for analyzing the impact of high-proportion distributed photovoltaic access to the distribution network that considers correlations to improve the operational reliability of the distribution network.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A method for analyzing the impact of high-proportion distributed photovoltaic (PV) grid integration considering correlations includes the following steps:

[0007] Based on the basic data of high-proportion distributed photovoltaic power to be connected, a probability density function of high-proportion distributed photovoltaic power is constructed.

[0008] Calculate the Spearman correlation coefficients among the power variables of high-proportion distributed photovoltaic power, and construct the first Spearman correlation matrix;

[0009] Using a Latin hypercube sampling method based on a genetic algorithm, and based on the probability density function, a high-proportion distributed photovoltaic power sampling matrix (PLHS) is obtained. M×NThe matrix PLHS M×N The error between the second Spearman correlation matrix and the first Spearman correlation matrix is ​​less than a set threshold, where M is the high-proportion distributed photovoltaic power variable and N is the number of samplings;

[0010] Data from the distribution network is acquired based on the high-proportion distributed photovoltaic power sampling matrix PLHS. M×N The electrical parameters of a high proportion of distributed photovoltaic (PV) power grid are calculated, and the impact analysis results of the high proportion of distributed PV power grid connection are obtained based on the electrical parameters.

[0011] Furthermore, the probability density function is expressed as:

[0012]

[0013] In the formula: bw is the bandwidth, P i For distributed photovoltaic power variables, P ij The original data for distributed photovoltaic power is given by n, where n is the number of original data samples and d(·) is the kernel function.

[0014] Furthermore, the bandwidth bw is selected based on minimizing the difference between the kernel distribution function and the empirical distribution function.

[0015] Furthermore, the Latin hypercube sampling method based on genetic algorithms specifically includes:

[0016] M random integer sequences from 1 to N are generated to form matrix I. M×N ;

[0017] Matrix I is optimized cyclically using a genetic algorithm. M×N The optimal matrix I is obtained. M×N,best The optimal matrix I M×N,best The third Spearman correlation matrix SP′ M×M,m The first Spearman correlation matrix SP M×M The error is less than the set threshold;

[0018] Based on the probability density function, the optimal matrix I M×N,best Inverse transformation yields the corresponding high-proportion distributed photovoltaic power sampling matrix PLHS M×N .

[0019] Furthermore, during the cyclic optimization process, the best individual obtained in the previous generation is retained in the next generation.

[0020] Furthermore, the fitness function of the iterative optimization process is set as follows:

[0021]

[0022] In the formula: S i For SP M×M The i-th data, S′ i,m SP′ of the m-th integer matrix M×M,m The i-th data.

[0023] Furthermore, the optimal matrix I M×N,best Inverse transformation yields the corresponding high-proportion distributed photovoltaic power sampling matrix PLHS M×N Specifically:

[0024] The optimal matrix I M×N,best All elements in the matrix are transformed as follows to obtain the numerical matrix F. M×N,best :

[0025]

[0026] In the formula: I j For I M×N,best The j-th data, F j For numerical matrix F M×N,best The j-th data;

[0027] Constructing an inverse distribution function based on the probability density function, for F M×N,best The distributed photovoltaic power value is obtained by transforming each row of data using the corresponding inverse distribution function, and a high-proportion distributed photovoltaic power sampling matrix PLHS is constructed. M×N .

[0028] Furthermore, regarding the high-proportion distributed photovoltaic power sampling matrix PLHS M×N For each column, the Newton-Raphson algorithm is used to perform power flow calculations to obtain the electrical parameters.

[0029] Furthermore, the electrical parameters include power flow distribution, line loss, system network loss, voltage deviation, voltage fluctuation, and harmonic current.

[0030] This invention also provides a system for analyzing the impact of high-proportion distributed photovoltaic (PV) grid integration on distribution networks, considering correlations, comprising:

[0031] The probability distribution construction module is used to construct the probability density function of high-proportion distributed photovoltaic based on the basic data of the high-proportion distributed photovoltaic to be connected;

[0032] The measured correlation matrix calculation module is used to calculate the Spearman correlation coefficient between high-proportion distributed photovoltaic power variables and construct the first Spearman correlation matrix.

[0033] The sampling matrix construction module is used to obtain a high-proportion distributed photovoltaic power sampling matrix (PLHS) based on the probability density function using a Latin hypercube sampling method based on a genetic algorithm. M×N The matrix PLHS M×N The error between the second Spearman correlation matrix and the first Spearman correlation matrix is ​​less than a set threshold, where M is the high-proportion distributed photovoltaic power variable and N is the number of samplings;

[0034] The impact analysis module is used to acquire distribution network data based on the high-proportion distributed photovoltaic power sampling matrix PLHS. M×N The electrical parameters of a high proportion of distributed photovoltaic (PV) power grid are calculated, and the impact analysis results of the high proportion of distributed PV power grid connection are obtained based on the electrical parameters.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. This invention considers the Spearman correlation coefficient between power variables of high-proportion distributed photovoltaic power, which can accurately analyze the impact of high-proportion distributed photovoltaic access on power flow distribution, network loss, voltage deviation, voltage fluctuation and harmonic current of distribution network. It has high reliability and can improve the operational reliability of distribution network.

[0037] 2. This invention uses a Latin hypercube sampling method based on genetic algorithm to obtain the sampling matrix, ensuring the consistency between the sampled data and the measured data, and further improving the reliability of the analysis of the influence of each indicator. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the process of the present invention;

[0039] Figure 2 This is an improved IEEE 33-node system used in embodiments of the present invention;

[0040] Figure 3 This is a power flow probability density diagram of branch 1-2 under different distributed photovoltaic capacities in an embodiment of the present invention;

[0041] Figure 4 This is a power flow probability density diagram of branch 15-16 under different distributed photovoltaic capacities in an embodiment of the present invention;

[0042] Figure 5 These are the expected loss curves for each branch in an embodiment of the present invention;

[0043] Figure 6 The curves showing the expected voltage fluctuations at each node in this embodiment of the invention are shown.

[0044] Figure 7This refers to the harmonic current value introduced by the distributed photovoltaic system at node 18 in this embodiment of the invention. Detailed Implementation

[0045] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0046] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0047] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0048] High-proportion distributed photovoltaic (PV) grid connections have a certain impact on various electrical indicators of the distribution network, including power flow distribution, network losses, voltage deviation, voltage fluctuations, and harmonic currents. Against this backdrop, it is crucial to rationally and accurately analyze the impact of distributed PV grid connections on the distribution network and understand the key factors that need to be considered to ensure the normal operation of the distribution network by examining the changes in various electrical indicators after distributed PV grid connection. Based on the above considerations, this invention provides a method for analyzing the impact of high-proportion distributed PV grid connections on distribution networks that considers correlations. Figure 1 As shown, the method includes the following steps:

[0049] Step S1: Based on the basic data of the high-proportion distributed photovoltaic power generation to be connected, construct the probability density function of the high-proportion distributed photovoltaic power generation.

[0050] Step S2: Calculate the Spearman correlation coefficients among the power variables of high-proportion distributed photovoltaic power, and construct the first Spearman correlation matrix;

[0051] Step S3: Using the Latin Hypercube Sampling (LHS) method based on a genetic algorithm, a high-proportion distributed photovoltaic power sampling matrix PLHS is obtained based on the probability density function. M×N The matrix PLHS M×N The error between the second Spearman correlation matrix and the first Spearman correlation matrix is ​​less than a set threshold, where M is the high-proportion distributed photovoltaic power variable and N is the number of samplings;

[0052] Step S4: Obtain distribution network data based on the high-proportion distributed photovoltaic power sampling matrix PLHS. M×N The electrical parameters of a high proportion of distributed photovoltaic (PV) power grid are calculated, and the impact analysis results of the high proportion of distributed PV power grid connection are obtained based on the electrical parameters.

[0053] In step S1, the probability density function is calculated based on the kernel density estimation method and can be expressed as:

[0054]

[0055] In the formula: bw is the bandwidth, P i For distributed photovoltaic power variables, P ij The original data for distributed photovoltaic power is given by n, where n is the number of original data samples and d(·) is the kernel function.

[0056] Kernel density estimation is a nonparametric estimation method that does not require pre-setting the distribution of the original data. It discovers patterns in the distribution using existing data, making it relatively easy to fit the data. The kernel distribution function F... Pi (P i ) can be achieved by f Pi (P i The bandwidth (bw) is obtained by integration, and its magnitude largely determines the accuracy of fitting the original data. Typically, a suitable bandwidth can be selected by minimizing the difference between the kernel distribution function and the empirical distribution function. If this difference is large, it indicates that the selected bandwidth is not suitable, and the resulting kernel distribution function cannot fit the original data well. This will also have a certain impact on the subsequent probabilistic power flow calculation results of the distribution network. Therefore, further adjustment of the bandwidth is necessary.

[0057] In step S2, the formula for calculating the Spearman correlation coefficient is:

[0058]

[0059] In the formula: P1 and P2 are distributed photovoltaic power variables. and Let S(P1,P2) be the distribution function, and ρ be the correlation coefficient function. If the absolute value of S(P1,P2) is close to 1, it indicates a very strong and strict monotonicity between P1 and P2; if the value is close to 0, it indicates no significant positive or negative correlation between P1 and P2. If S(P1,P2) is positive, it indicates a positive correlation between P1 and P2; if S(P1,P2) is negative, it indicates a negative correlation between P1 and P2.

[0060] Furthermore, the Spearman correlation coefficient is not related to the probability distribution function of distributed photovoltaic power. Therefore, the Spearman correlation coefficient can be used to describe the correlation between distributed photovoltaic power.

[0061] After calculating the Spearman correlation coefficients between each pair of distributed photovoltaic power variables, the first Spearman correlation matrix SP can be constructed. M×M .

[0062] In step S3, LHS sampling is used to obtain data for impact analysis. The purpose of LHS sampling is to ensure that the sampling points are not locally concentrated in a certain distribution interval. The distribution function ranges from [0,1], and this interval is divided into [0,n1], [n1,n2], ..., [n...]. N-1 The system first divides the system into N equal intervals, and then randomly selects a value from each interval to obtain N random values. The inverse distribution function of each random value is then calculated to obtain the corresponding N distributed photovoltaic power values. Since there are M distributed photovoltaic power variables, and the data sampled from each variable is one row of a sampling matrix, an M x N sampling matrix PLHS can be obtained. M×N .

[0063] There is a correlation in the distributed photovoltaic (PV) power among the M distributed PV power plants; therefore, the Spearman correlation coefficient needs to be considered in the obtained M sets of data. Since the corresponding distribution function value decreases as the distributed PV power decreases, the Spearman correlation coefficient between distribution functions is equal to the Spearman correlation coefficient between distributed PV power. Therefore, to ensure that the Spearman correlation coefficient matrix of the sampled distributed PV power data matches the Spearman correlation matrix of the measured distributed PV power, the Spearman correlation coefficient matrix needs to be considered. M×M To make them relatively close, the correlation matrix of the M groups of distribution functions obtained from sampling should be similar to that of SP. M×M They are quite close. To achieve the above goal, M sets of integer sequences from 1 to N can be generated to form matrix I. M×N Make its Spearman correlation matrix SP′ M×M Basic and SP M×M Similarly, it can be obtained through optimization using a genetic algorithm.

[0064] The reason for using a genetic algorithm to optimize the LHS sampling process is that the randomly generated I M×N SP′ of the matrix M×M Matrix is ​​often associated with SP M×M The differences are significant; by using a genetic algorithm for iterative optimization, SP′ can be improved. M×M With SP M×MThe gap between them is reduced to within a certain range. In each iteration of optimization, when generating L... ga Personal I M×N After obtaining the matrix, the integer matrices with smaller fitness functions are selected to form the parent and mother integer matrices. In this invention, the fitness function is set as follows:

[0065]

[0066] In the formula: S i For SP M×M The i-th data, S′ i,m SP′ of the m-th integer matrix M×M,m The i-th data. During the optimization process, fitness(I) needs to be optimized. m The value gradually decreases, mainly by modifying I. M×N This is achieved by using the values ​​of the elements in the matrix.

[0067] Minimum fitness (I) m The integer matrix corresponding to the selected matrix is ​​the optimal matrix. The optimal matrix is ​​used as the parent integer matrix, and the remaining selected matrices are used as the parent integer matrices. A row of data from the optimal matrix and the parent integer matrices is randomly swapped to form the offspring matrix. It is important to note that to prevent the optimal individual from being destroyed during the swapping process, it needs to be preserved in the offspring. Then, the elements in the offspring matrix are mutated to obtain a richer integer matrix. For each row of the integer matrix, a value between 0 and 1 is randomly generated. If this value is greater than a set mutation rate, the two elements in that row are swapped. During this process, the optimal matrix cannot be mutated to ensure it is saved to the next integer matrix population.

[0068] The optimal matrix I is obtained through the GA algorithm. M×N,best Then, the optimal matrix I M×N,best All elements in the matrix are transformed as follows to obtain the numerical matrix F. M×N,best :

[0069]

[0070] In the formula: I j For I M×N,best The j-th data, F j For numerical matrix F M×N,best The j-th data;

[0071] Constructing an inverse distribution function based on the probability density function, for F M×N,best The distributed photovoltaic power value is obtained by transforming each row of data using the corresponding inverse distribution function, and a high-proportion distributed photovoltaic power sampling matrix PLHS is constructed. M×N .

[0072] In step S4, the Newton-Raphson algorithm is used to calculate the power flow and obtain electrical parameters, including the voltage amplitude and phase angle of each node and the power flow of each branch. The expected value, standard deviation and probability distribution of the power flow and voltage (including amplitude and phase angle) of the high-proportion distributed photovoltaic power grid are calculated. Based on the calculation results of the power flow distribution, line loss, system network loss, voltage deviation, voltage fluctuation and harmonic current of the distribution network, the relevant impacts of the high-proportion distributed photovoltaic power grid are obtained.

[0073] In addition, after obtaining the corresponding calculation results, they can be compared with the corresponding indicators of historical data to obtain the error of the model. The specific calculation method is as follows:

[0074]

[0075] In the formula: ε is the relative error, para′ and para are the electrical indicators obtained from the sampled data and historical data, respectively, and dig is the statistical indicator, including the expected value and standard deviation.

[0076] After obtaining the relevant impact analysis results, corresponding risk reduction measures can be generated based on the impact analysis results to improve the reliability of the distribution network operation after a high proportion of distributed photovoltaic power is connected to the distribution network.

[0077] Example

[0078] In this embodiment, with Figure 2 Simulations were performed as test cases. Figure 2 In this scenario, the active power demand of the load is 3715kW, and the reactive power demand is 2300kvar. Distributed photovoltaic (PV) power sources are connected to nodes 18 and 22 respectively. Correlation coefficient analysis is performed on the output of the two sources, and probabilistic power flow is calculated based on LHS (Low Power Flow) to analyze how the electrical indicators of the distribution network change with the change of distributed PV capacity.

[0079] Calculations show that the Spearman coefficient matrix of the measured distributed photovoltaic power data for the two photovoltaic power plants is as follows:

[0080] Sampling was performed using the GA-optimized LHS, with 1000 sampling iterations. After processing the integer matrix, distributed photovoltaic power data was obtained through inverse distribution function transformation. The Spearman correlation coefficient matrices of these two sets of distributed photovoltaic power data are as follows:

[0081] Therefore, it can be seen that the Spearman correlation coefficient matrix of the distributed photovoltaic power data obtained by LHS sampling is basically similar to the Spearman correlation coefficient matrix of the measured data, with a small error.

[0082] Based on the power flow calculation results, the changes in various indicators such as power flow distribution, network loss, voltage deviation, voltage fluctuation and harmonic current in this embodiment are calculated, and the impact of high proportion of distributed photovoltaic access on the distribution network is analyzed.

[0083] Regarding power flow distribution, the power flow probability density functions for branches 1-2 and 15-16 under different distributed photovoltaic access capacities are as follows: Figure 3 and Figure 4 As shown.

[0084] Depend on Figure 3 It can be seen that after distributed photovoltaic (PV) grid connection, the active and reactive power on lines far from the grid connection point will decrease compared to when distributed PV is not connected. When the distributed PV capacity does not exceed 2500kW, the reduction in power flow in branch 1-2 gradually increases with the increase in distributed PV capacity.

[0085] Depend on Figure 4 It can be seen that when the distributed photovoltaic power is large enough, the power flow of the branch near the distributed photovoltaic access point will change direction. Moreover, as the distributed photovoltaic capacity increases, the reverse power flow tends to increase. In fact, the amplitude of active power and reactive power may even exceed the amplitude when not connected.

[0086] Regarding network losses, the expected loss curves for each branch of the distribution network are as follows: (The curves show the expected losses for each branch with and without distributed photovoltaic power sources, respectively.) Figure 5 As shown. By Figure 5 It can be seen that after the distributed photovoltaic (PV) system is connected, the power flow in branches farther from the PV connection point will decrease, thus reducing the losses on those branches. For example, the expected loss on branch 2-3 decreases from 51.8kW to 39.3kW. However, the power flow in branches closer to the PV connection point will reverse, and the power flow amplitude may increase, leading to increased losses on those branches. For instance, the expected losses on branches 13-14, 14-15, 15-16, 16-17, 17-18, and 21-22 are higher than when no PV system was connected.

[0087] Regarding voltage offset, the calculation results of the probability that the voltage amplitude of some nodes exceeds the specified range under different distributed photovoltaic capacities are shown in Table 1.

[0088] Table 1

[0089]

[0090]

[0091] As shown in Table 1, when the distributed photovoltaic capacity is less than 2500kW, the probability of some nodes having voltage amplitudes exceeding the specified range decreases as the distributed photovoltaic capacity increases. For example, the probability of node 16 decreasing from 53.2% to 27.3% when the distributed photovoltaic capacity increases from 1000kW to 2500kW.

[0092] Regarding voltage fluctuations, the expected value curves of voltage fluctuations at each node are as follows: Figure 6 As shown in the figure. Analysis shows that the closer the node is to the distributed photovoltaic access point, the greater the expected value of the voltage fluctuation; the expected value of the voltage fluctuation of each node increases with the increase of the distributed photovoltaic capacity.

[0093] Regarding harmonic currents, when the connected distributed photovoltaic capacity is 2500kW, the values ​​of harmonic currents of different orders introduced by the distributed photovoltaic at node 18 are as follows: Figure 7 As shown. By Figure 7 It can be seen that after distributed photovoltaic (PV) grid connection, the distribution network will generate harmonic currents. The odd-order harmonic current values ​​are generally larger than the even-order harmonic current values, with the third and fifth harmonics having a greater impact on the distribution network. Therefore, when considering the impact of distributed PV grid connection on the distribution network, harmonic currents cannot be ignored, and it is necessary to prevent them from exceeding a certain limit.

[0094] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A method for analyzing the impact of high penetration of distributed photovoltaic access to a power distribution network taking into account correlations, characterized in that, Includes the following steps: Based on the basic data of high-proportion distributed photovoltaic power to be connected, the probability density function of high-proportion distributed photovoltaic power is constructed by kernel density estimation method. Calculate the Spearman correlation coefficients among the power variables of high-proportion distributed photovoltaic power, and construct the first Spearman correlation matrix; Using a Latin hypercube sampling method based on a genetic algorithm, and based on the probability density function, a high-proportion distributed photovoltaic power sampling matrix is ​​obtained. The matrix The error between the second Spearman correlation matrix and the first Spearman correlation matrix is ​​less than a set threshold, wherein, M For high proportion of distributed photovoltaic power variables, N Number of samples; Acquire distribution network data based on the high-proportion distributed photovoltaic power sampling matrix. Calculate the electrical parameters after a high proportion of distributed photovoltaic power is connected to the distribution network, and obtain the impact analysis results of the high proportion of distributed photovoltaic power being connected to the distribution network based on the electrical parameters; The Latin hypercube sampling method based on genetic algorithms specifically includes: Randomly generated M The group consists of 1 to N integer sequence, the matrix ; The matrix is ​​optimized iteratively using a genetic algorithm. To obtain the optimal matrix The optimal matrix The third Spearman correlation matrix Correlation matrix with the first Spearman The error is less than the set threshold; based on the probability density function, a best matrix inverse transform into a corresponding high proportion distributed photovoltaic power sample matrix ; During the iterative optimization process, the best individual obtained in the previous generation is retained in the next generation; the fitness function of the iterative optimization process is set as follows: In the formula: for The i One data point, For the first m an integer matrix The i One data point; Said optimal matrix Inverse transform into corresponding high proportion distributed photovoltaic power sampling matrix Specifically: transforming all elements in the optimal matrix to obtain the numerical matrix : In the formula: for The j One data point, Numerical matrix The j One data point; Based on the probability density function, the inverse distribution function is constructed, and the distributed photovoltaic power value is obtained by transforming each row of data using the corresponding inverse distribution function. The high-proportion distributed photovoltaic power sampling matrix is constructed. .

2. The method for analyzing the impact of high penetration of distributed photovoltaic access power distribution network considering correlation according to claim 1, characterized in that, The probability density function is expressed as: In the formula: For bandwidth, For distributed photovoltaic power variables, This is the raw data for distributed photovoltaic power. n The number of original data samples. This is a kernel function.

3. The method for analyzing the impact of high penetration of distributed photovoltaic access power distribution network considering correlation according to claim 2, characterized in that, The bandwidth Based on selecting by minimizing the difference between the kernel distribution function and the empirical distribution function.

4. The method for analyzing the impact of high penetration of distributed photovoltaic access power distribution network considering correlation according to claim 1, characterized in that, For each column of the high proportion distributed photovoltaic power sampling matrix , the Newton-Raphson algorithm is used for power flow calculation to obtain the electrical parameters.

5. The correlation-aware high penetration distributed PV access impact on distribution network analysis method according to claim 1 or 4, characterized in that, The electrical parameters include power flow distribution, line loss, system network loss, voltage deviation, voltage fluctuation, and harmonic current.

6. A system for analyzing the impact of high penetration of distributed photovoltaic access to a power distribution network taking into account correlation, characterized in that, include: The probability distribution construction module is used to construct the probability density function of high-proportion distributed photovoltaic based on the basic data of the high-proportion distributed photovoltaic to be connected, using the kernel density estimation method. The measured correlation matrix calculation module is used to calculate the Spearman correlation coefficient between high-proportion distributed photovoltaic power variables and construct the first Spearman correlation matrix. The sampling matrix construction module is used to obtain a high-proportion distributed photovoltaic power sampling matrix based on the probability density function using a Latin hypercube sampling method based on a genetic algorithm. The matrix The error between the second Spearman correlation matrix and the first Spearman correlation matrix is ​​less than a set threshold, wherein, M For high proportion of distributed photovoltaic power variables, N Number of samples; The impact analysis module is used to acquire distribution network data based on the high-proportion distributed photovoltaic power sampling matrix. Calculate the electrical parameters after a high proportion of distributed photovoltaic power is connected to the distribution network, and obtain the impact analysis results of the high proportion of distributed photovoltaic power being connected to the distribution network based on the electrical parameters; The Latin hypercube sampling method based on genetic algorithms specifically includes: Randomly generated M The group consists of 1 to N integer sequence, the matrix ; Optimizing a matrix by genetic algorithm loop to obtain an optimal matrix The third Spearman correlation matrix of the optimal matrix has an error less than a set threshold value from the first Spearman correlation matrix ​ based on the probability density function, a best matrix inverse transform to a corresponding high proportion distributed photovoltaic power sample matrix ; During the iterative optimization process, the best individual obtained in the previous generation is retained in the next generation; the fitness function of the iterative optimization process is set as follows: In the formula: for The i One data point, For the first m an integer matrix The i One data point; Said optimal matrix Inverse transform into corresponding high proportion distributed photovoltaic power sampling matrix Specifically: transforming all elements in the optimal matrix to obtain the numerical matrix : wherein: is the j th data of the numerical matrix is j the th Constructing an inverse distribution function based on the probability density function, for The distributed photovoltaic power value is obtained by transforming each row of data using the corresponding inverse distribution function, and a high-proportion distributed photovoltaic power sampling matrix is ​​constructed. .

Images