Method for quantifying responsibility of multi-harmonic source based on reconstructed data processing

Abstract

The application provides a multi-harmonic source responsibility quantification method based on reconstructed data processing. For a target harmonic source on a feeder, the least square method is used to fit an initial equivalent harmonic impedance value corresponding to the target harmonic source, and the initial value is used to reconstruct the power grid background harmonic voltage data. The data is sorted and segmented to reduce the fluctuation, and the independent random vector method is used to calculate the harmonic impedance in each segment. The impedance initial value and the difference between the impedance of each segment are used as the criteria to construct a target function, and the optimal initial value meeting the target function is used as the harmonic impedance estimation value of the harmonic feeder to quantify the harmonic responsibility of the feeder. The application can provide accurate quantification results when the power grid background harmonic has strong fluctuation.

Description

A Multi-Harmonic Source Responsibility Quantification Method Based on Reconstructed Data Processing Technical Field

[0001] This invention relates to the field of centralized multi-harmonic source responsibility quantification technology, and in particular to a multi-harmonic source responsibility quantification method based on reconstructed data processing. Background Technology

[0002] Due to the widespread use of power electronic equipment in power systems, there are numerous harmonic sources in the power grid, resulting in extremely high harmonic content and severe harmonic pollution. Accurately distinguishing the individual harmonic contributions of each harmonic source on the bus of concern is of great significance for harmonic control and quantifying harmonic responsibility.

[0003] Existing methods for quantifying the responsibility of multiple harmonic sources typically assume that background harmonics remain stable. However, in modern power systems, distribution networks are highly electronic, and projects such as electric vehicle charging stations and urban rail transit involve a large number of nonlinear loads. In addition, the grid connection of new energy equipment such as wind power and photovoltaics causes severe fluctuations in background harmonics in the power grid. As a result, there is a large error between the harmonic contribution estimated by existing methods and the true value. Summary of the Invention

[0004] This invention addresses the problem of significant errors between the harmonic contribution estimated by existing methods and the true value. It proposes a multi-harmonic source responsibility quantification method based on reconstructed data processing, which can provide accurate quantification results when there are strong fluctuations in the background harmonics of the power grid.

[0005] To achieve the above objectives, the following technical solution is proposed:

[0006] The multi-harmonic source responsibility quantization method based on reconstructed data processing includes the following steps:

[0007] (A) Based on the Norton equivalent circuit of the multi-harmonic source at a specific harmonic frequency h, establish the equivalent circuit equation and use the least squares method to fit the initial value of the equivalent harmonic impedance of the feeder of interest.

[0008] (B) Substitute the initial value of the equivalent harmonic impedance obtained in step (A) into the equivalent circuit equation to reconstruct the background harmonic voltage data, and sort and segment the background harmonic voltage data to obtain several data segments. The corresponding measurement data and the background harmonic voltage data are processed in the same way.

[0009] (C) Calculate the equivalent harmonic impedance of each data segment using the random independent vector method, thereby obtaining the harmonic impedance of each data segment;

[0010] (D) Construct an objective function based on minimizing the difference between the calculated impedance values ​​of each data segment obtained in step (C) and the initial impedance values. Iterate to find the optimal initial value that satisfies the objective function as the estimated harmonic impedance value of the feeder of concern. Finally, use the estimated equivalent harmonic impedance value of the feeder of concern and the vector projection of the harmonic current on the harmonic voltage of the bus of concern to quantify the contribution of the harmonic source.

[0011] For the target harmonic source on the feeder of concern, this invention first uses the least squares method to fit the measurement data to obtain the initial value of the equivalent harmonic impedance corresponding to the target harmonic source, and then uses this initial value to reconstruct the background harmonic voltage data of the power grid. This data is sorted and segmented to reduce fluctuations, and the harmonic impedance of each segment is calculated using the independent random vector method. An objective function is constructed based on minimizing the difference between the initial impedance value and the calculated impedance of each segment. The optimal initial value satisfying the objective function is iteratively used as the harmonic impedance estimate of the feeder of concern, thereby quantifying the harmonic responsibility of the feeder. This invention, based on a multi-harmonic source responsibility quantification method using reconstructed data processing, can reconstruct the background harmonic voltage using measurement data. By processing the reconstructed background harmonic voltage data, the impact of its fluctuations on the algorithm is reduced, resulting in more robust and accurate estimation results. This invention uses minimizing the difference between the calculated impedance and the initial impedance value as the criterion, and iteratively converges to the optimal impedance value as the final result, regardless of the initial impedance value setting.

[0012] Preferably, step (A) specifically includes the following steps:

[0013] (A1) When multiple feeders are connected to the busbar of the node of interest, the Norton equivalent circuit with multiple harmonic sources is used for analysis. The system side and each feeder are equivalent to the parallel connection of the corresponding harmonic current source and harmonic impedance. Let Z0 and Z k (k = 1, 2, ..., n) represent the harmonic impedances of the system side and each feeder, respectively; and These are the harmonic currents emitted by the harmonic sources on the system side and each feeder, respectively. The harmonic current of the feeder; by the superposition theorem, we are concerned with the harmonic voltage at the bus. It is the sum of the harmonic voltages generated at the PCC when each feeder harmonic source acts individually, i.e.

[0014]

[0015] in It is the harmonic voltage generated at PCC by the harmonic source on the k-th feeder; It is the harmonic voltage generated at the PCC by the system-side harmonic source, i.e., the background harmonic voltage, denoted as...

[0016]

[0017] Z pcc-0 Z is the equivalent impedance excluding system-side harmonic impedance; pcc-k The equivalent harmonic impedance excluding the harmonic source of feeder k, i.e., the parallel connection of the harmonic impedances of other branches besides feeder k:

[0018]

[0019] Harmonic contribution index HC of feeder k pcc-k It can be determined by the vector projection of the harmonic voltage it generates at the PCC point onto the total harmonic voltage at the PCC point:

[0020]

[0021] The determination of harmonic contribution depends on the harmonic impedance Z. pcc-k Accurate estimation;

[0022] (A2) in The target harmonic source feeder i is separated, and the remaining feeder harmonic sources are combined with the background harmonic voltage term.

[0023]

[0024] Assuming the harmonic impedance remains constant, the following equation is obtained for the j-th sampled data:

[0025]

[0026] Where ε j For random error, Consider it as a linear regression model:

[0027] Y = Xθ + ε

[0028] in

[0029]

[0030]

[0031]

[0032] The regression coefficient θ is fitted using the least squares method:

[0033]

[0034] Since there is no other known information besides the harmonic measurement data, the initial value of the harmonic impedance is first fitted using the least squares method.

[0035] Preferably, step (B) specifically includes the following steps:

[0036] (B1) in Divide both ends by phase angle

[0037]

[0038] Where α is and The phase angle between them; α0 is and Phase angle between; separation The real and imaginary parts can be obtained as follows:

[0039]

[0040] Z pc-c and Z pcc-iy Let be the real and imaginary parts of the harmonic impedance, respectively. get:

[0041]

[0042] (B2) Substitute the initial value of the harmonic impedance obtained in (A1) into... Reconstructing background harmonic voltage data E x(y) ; background harmonic data E x(y) Sort and segment the data from smallest to largest, and the corresponding measurement data U x(y) , Also according to E x(y) Arrange them in order and divide them into segments.

[0043] Preferably, step (C) specifically includes the following steps:

[0044] (C1) In the calculation of the real part, assuming the total number of samples is N, the number of segments is P, and the sample size of each sub-segment is n = N / P, the harmonic voltage equation for each data segment can be obtained:

[0045]

[0046] The subscript k represents the equation of the k-th data segment, E x It is the initial impedance value fitted in step (A2). Substitution The inverse solution is obtained by further segmentation, U x and The samples need to be in accordance with E x Arrange them in the same order and divide them into segments;

[0047] (C2) The formula for calculating the harmonic impedance in each data segment is:

[0048] The harmonic impedance of each data segment can be calculated using this formula.

[0049] Preferably, the derivation process of the calculation formula for the harmonic impedance in each data segment is as follows: The average value is:

[0050]

[0051] Multiply both ends simultaneously Taking the average again yields:

[0052]

[0053] E x This includes background harmonics emitted from the system side and harmonics emitted from other feeder harmonic sources besides feeder i, as well as the statistical independence between different harmonic sources, E. x and If they are independent or only weakly correlated, their covariance is 0, i.e.

[0054]

[0055] Will Substitution The formula for calculating the harmonic impedance in each data segment can be obtained as follows:

[0056]

[0057] Preferably, step (D) specifically includes the following steps:

[0058] (D1) Define the difference function ΔZ between the calculated impedance and the initial impedance value for each data segment. The calculation formula is as follows:

[0059]

[0060] P represents the number of sub-segments, each input impedance initial value. This will result in a difference function ΔZ. x ΔZ x It is the initial value of impedance. The function; when ΔZ x The minimum value indicates that the difference between the initial impedance value and the calculated impedance value of each data segment is the smallest. This initial impedance value is the optimal initial impedance value, and the objective function is:

[0061] (D2) Define the impedance change step size δ to iterate the initial impedance value, and determine the direction of iteration by calculating γ.

[0062]

[0063] The specific direction of the initial impedance iteration is as follows:

[0064]

[0065] (D3) Repeat steps (B1) to (D2) until the difference function ΔZ is obtained. x No further increases were made, and ΔZ was verified within a range of 10 times the initial impedance value. x If it is the minimum value, then the initial impedance value corresponding to the minimum value of the difference function is what the algorithm seeks.

[0066]

[0067] If it is not the minimum value, the initial value of the iterative impedance is recalculated. After calculating the harmonic impedance, it is then calculated using the formula... Calculate the harmonic contribution of the harmonic source of feeder i.

[0068] The beneficial effects of this invention are:

[0069] 1. The present invention is a multi-harmonic source responsibility quantization method based on reconstructed data processing. It can reconstruct the background harmonic voltage using measurement data, and by processing the reconstructed background harmonic voltage data, it reduces the impact of its fluctuations on the algorithm, resulting in more robust and accurate estimation results.

[0070] 2. This invention uses minimizing the difference between the calculated impedance and the initial impedance value as the criterion, and uses the optimal impedance as the final result through iteration. Regardless of the initial impedance value setting, the algorithm can iteratively converge to the optimal initial impedance value. Attached Figure Description

[0071] Figure 1 is a schematic diagram of the process of this invention;

[0072] Figure 2 shows the topology of the centralized multi-harmonic source in step (A1);

[0073] Figure 3 shows the Norton equivalent circuit model diagram for step (A1) and Example 1;

[0074] Figure 4 shows the harmonic voltage phase diagram in step (A1);

[0075] Figure 5 shows a comparison of the data processing results in step (B2). (a) Original data distribution; (b) Data distribution after processing;

[0076] Figure 6 shows the curves of the initial impedance value and the difference function in the embodiment. (a) Real part curve; (b) Imaginary part curve;

[0077] Figure 7 is a line graph showing the mean square error of the harmonic contribution of each feeder calculated by each method when the background harmonic fluctuation increases in the embodiment. (a) Feeder 1; (b) Feeder 2; (c) Feeder 3; (d) Feeder 4;

[0078] Figure 8 is a line graph showing the mean square error of the harmonic contribution of each feeder calculated by each method when the system impedance changes in the embodiment. (a) Feeder 1; (b) Feeder 2; (c) Feeder 3; (d) Feeder 4; Detailed Implementation

[0079] Example:

[0080] This embodiment proposes a multi-harmonic source responsibility quantification method based on reconstructed data processing, referring to Figure 1, including the following steps:

[0081] (A) Based on the Norton equivalent circuit of a multi-harmonic source at a certain harmonic frequency, establish the equivalent circuit equation and use the least squares method to fit the initial value of the equivalent harmonic impedance of the feeder of interest.

[0082] The Norton equivalent equation establishment and the fitting of the initial value of the harmonic impedance in step (A) are carried out according to the following steps:

[0083] (A1) As shown in Figures 2 and 4, the scenario where multiple feeders are connected to the node bus of interest, i.e., the Point of Common Coupling (PCC), is called a centralized multi-harmonic source system. It can typically be analyzed using the Norton equivalent model shown in Figure 3, where the system side and each feeder are equivalently represented as the parallel connection of the corresponding harmonic current source and harmonic impedance. Let Z0 and Z... k (k = 1, 2, ..., n) represent the harmonic impedances of the system side and each feeder, respectively; and These are the harmonic currents emitted by the harmonic sources on the system side and each feeder, respectively. This refers to the harmonic current of the feeder. By the superposition theorem, we are concerned with the harmonic voltage at the bus. It is the sum of the harmonic voltages generated at the PCC when each feeder harmonic source acts individually, i.e.

[0084]

[0085] in It is the harmonic voltage generated at PCC by the harmonic source on the k-th feeder; It is the harmonic voltage generated at the PCC by the system-side harmonic source, i.e., the background harmonic voltage, which can be expressed as:

[0086]

[0087] Z pcc-0 Z is the equivalent impedance excluding system-side harmonic impedance; pcc-k The equivalent harmonic impedance excluding the harmonic source of feeder k, i.e., the parallel connection of the harmonic impedances of other branches besides feeder k:

[0088]

[0089] As shown in Figure 3, the harmonic contribution index HC of feeder k pcc-k It can be determined by the vector projection of the harmonic voltage it generates at the PCC point onto the total harmonic voltage at the PCC point:

[0090]

[0091] As can be seen from equation (4), the determination of harmonic contribution depends on the harmonic impedance Z. pcc-k An accurate estimate.

[0092] (A2) Separate the target harmonic source feed line i from equation (1), and combine the remaining feed line harmonic sources with the background harmonic voltage term.

[0093]

[0094] Assuming the harmonic impedance remains constant, the following equation can be obtained for the j-th set of sampled data:

[0095]

[0096] Where ε j As for random error, equation (5) can be regarded as a linear regression model:

[0097] Y = Xθ + ε (7)

[0098] in

[0099]

[0100]

[0101]

[0102] The regression coefficient θ can be fitted using the least squares method:

[0103]

[0104] Regression models typically require constant coefficients, where harmonic impedance can be considered constant under steady-state conditions, while The harmonic currents emitted by other feeder harmonic sources and the background harmonics emitted by the system side are difficult to keep constant. Therefore, when the background harmonics fluctuate greatly, directly using Equation (8) to calculate the harmonic impedance will result in a large deviation from the true value.

[0105] Since there is no other known information besides the harmonic measurement data, the harmonic impedance can be fitted using the least squares method as an initial value. The initial impedance value of this invention is denoted as follows: The initial impedance value is then corrected in subsequent steps to obtain the desired value.

[0106] (B) Substitute the initial harmonic impedance value obtained in step (A) into the equivalent circuit equation to reconstruct the background harmonic voltage data, and sort and segment this background harmonic voltage data. The corresponding measurement data is also processed in the same way as the background harmonics. Step (B), which reconstructs the background harmonic voltage data using the initial harmonic impedance value and performs sorting and segmentation processing, is carried out according to the following steps:

[0107] (B1) Divide both sides of equation (5) by phase angle

[0108]

[0109] Where α is and The phase angle between them; α0 is and The phase angle between them; the real and imaginary parts of the separated equation (9) can be obtained as follows:

[0110]

[0111] in and Z pcc-iyLet be the real and imaginary parts of the harmonic impedance, respectively. For simplicity, let be... The above equation can be simplified to:

[0112]

[0113] (B2) Substitute the initial value of the harmonic impedance obtained in (B1) into equation (11) to reconstruct the background harmonic voltage data E. x(y) Background harmonic data E x(y) Sort and segment the data from smallest to largest, and the corresponding measurement data U x(y) , Also according to E x(y) The data is arranged and segmented in the correct order. As shown in Figure 5, through processing the reconstructed data, it can be found that in the sorted and segmented data, U... x(y) and The linear correlation is higher, and the background harmonic variance of each data segment is smaller than that of the source data. Therefore, using the sorted and segmented data to calculate the harmonic impedance results in less error and more accurate results.

[0114] (C) In each background harmonic sub-segment obtained in step (B), the equivalent harmonic impedance of the data segment is calculated using the random independent vector method, thereby obtaining the harmonic impedance of each data segment.

[0115] In step (C), the equivalent harmonic impedance of each data segment is calculated using the random independent vector method within each sub-segment of the background harmonics, thereby obtaining the harmonic impedance of each data segment. This is performed according to the following steps:

[0116] (C1) The above steps reconstruct the background harmonic data by fitting the initial impedance value, and reduce fluctuations by sorting and segmenting it. Taking the calculation of the real part as an example, assuming the total number of samples is N, the number of segments is P, and the sample size of each sub-segment is n = N / P, the harmonic voltage equation for each data segment can be obtained:

[0117]

[0118] In equation (12), the superscript k represents the equation for the k-th data segment. It should be noted that E x It is the initial impedance value fitted in step (A2). Substituting into equation (11) and solving inversely, and then dividing it into segments, U x and The samples need to be in accordance with E x Arrange them in the same order and divide them into segments.

[0119] Equation (2) (C2) yields the mean value:

[0120]

[0121] Multiply both sides of equation (13) by Taking the average again yields:

[0122]

[0123] E x This includes background harmonics emitted from the system side and harmonics emitted from other feeder harmonic sources besides feeder i, determined by the statistical independence between different harmonic sources, E x and They are independent or only weakly correlated, therefore their covariance is 0, i.e.

[0124]

[0125] Substituting equation (15) into equation (14) yields the formula for calculating the harmonic impedance in each data segment:

[0126]

[0127] (D) Construct an objective function based on minimizing the difference between the calculated impedance values ​​of each data segment obtained in step (C) and the initial impedance values. Iterate to find the optimal initial value that satisfies the objective function as the estimated harmonic impedance value of the feeder of interest. Finally, use the estimated equivalent harmonic impedance of the feeder of interest and the vector projection of the harmonic current onto the harmonic voltage of the bus of interest to quantify the contribution of the harmonic source.

[0128] In step (D), the objective function is defined, the initial impedance is iterated, and the harmonic contribution is solved. This is performed according to the following steps:

[0129] (D1) Define the difference function ΔZ between the calculated impedance and the initial impedance value for each data segment. The calculation formula is as follows:

[0130]

[0131] P represents the number of sub-segments, each input impedance initial value. This will result in a difference function ΔZ. x Therefore, ΔZ x It is the initial value of impedance. The function. When ΔZ x The minimum value indicates that the difference between the initial impedance value and the calculated impedance value of each data segment is the smallest. This initial impedance value is the optimal initial impedance value, and the objective function is:

[0132] (D2) Define the impedance change step size δ to iterate the initial impedance value, and determine the direction of iteration by calculating γ.

[0133]

[0134] The specific direction of the initial impedance iteration is as follows:

[0135]

[0136] (D3) Repeat steps (B1) to (D2) until the difference function ΔZ is obtained. x No further increases were made, and ΔZ was verified within a range of 10 times the initial impedance value. x If it is the minimum value, then the initial impedance value corresponding to the minimum value of the difference function is what the algorithm seeks.

[0137]

[0138] Conversely, if it is not the minimum value, the initial value of the iterative impedance is recalculated. Because ΔZ x The minimum value exists and is unique. The selection of the initial impedance value does not affect the final calculation result, but it does affect the computational complexity of the entire algorithm iteration process. This is because the goal of the proposed method is to find the value that makes the difference function ΔZ... x The proposed method converges to the minimum initial impedance value if this initial impedance value deviates significantly from the true value. x The computation time for the value will increase, while the minimum ΔZ x Since the initial impedance value is unique and converges to the minimum value after iteration, the selection of the initial impedance value itself does not affect the accuracy of the result.

[0139] After calculating the harmonic impedance, the harmonic contribution of the harmonic source of feeder i is calculated by equation (4).

[0140] The specific implementation process of this embodiment is as follows: Taking a 5th harmonic system with 4 feeders as an example, the Norton equivalent circuit shown in Figure 3 is built in Matlab, and the parameter settings are shown in Table 1.

[0141] Table 1 Simulation parameter settings in the embodiments

[0142]

[0143] The harmonic impedance of each feeder and the system side is equivalent to a series connection of a resistor and a reactance. Variance is used to measure the degree of data fluctuation. A random fluctuation with a mean of 0 and a variance of 1 is added to the harmonic current of each feeder. The coefficient K is used to adjust the background harmonic variance. A random fluctuation with a mean of 0 and a variance of 0.05K is superimposed on the harmonic current of the system side.

[0144] To further illustrate the calculation process of the method of this invention, taking the calculation of the harmonic contribution of feeder 1 as an example, it can be seen from equation (3) that Z pcc-1 The theoretical value of the harmonic impedance of feeders 2, 3, and 4 connected in parallel with the system side is 1.63 + j2.90 Ω. Let K = 2. The initial impedance value fitted using the least squares method is 1.48 + j1.93 Ω, and the optimal initial impedance value after iteration using the algorithm of this invention is 1.66 + j2.89 Ω. The difference function between the real and imaginary parts is calculated within a range of 10 times this initial value. The curves of the change between the initial impedance value and the difference function are shown in Figure 6. It can be seen that the difference function between the real and imaginary parts has a minimum value near the desired value. Then, the harmonic contribution HC of feeder 1 can be obtained from equation (4). pcc-1 It is 23.30%.

[0145] To analyze the computational accuracy of the proposed method and existing methods, we selected Method 1 (least squares method), Method 2 (covariance method), and Method 3 (the method proposed in this invention) for comparative analysis. The results are shown in Table 2.

[0146] Table 2 shows the results of calculating the harmonic contribution of each harmonic source using the present invention and other methods in the embodiments.

[0147]

[0148] To analyze the evaluation accuracy of the proposed method when background harmonic fluctuations increase, K was adjusted to 1, 2, 3, 4, and 5. The program was run 100 times, and the root mean square error (RMSE) of the harmonic contribution error for each feeder was calculated using the three methods described above. The results are shown in Figure 7.

[0149] To analyze the effectiveness of the proposed method when the system-side impedance and the user's harmonic impedance are close in magnitude, a coefficient m was set to adjust the magnitude of the system-side impedance. That is, the system-side parameters were set as follows: R = 2m, L = 0.002m; m was adjusted to 1, 3, and 5 respectively. The root mean square value of the harmonic contribution error of each feeder was calculated again using the above three methods. The results are shown in Figure 8.

Claims

1. A method for quantifying the responsibility of multiple harmonic sources based on reconstructed data processing, characterized in that, Includes the following steps: (A) Based on the Norton equivalent circuit of the multi-harmonic source at a specific harmonic frequency h, establish the equivalent circuit equation and fit the initial value of the equivalent harmonic impedance of the feeder of interest using the least squares method; (B) Substitute the initial value of the equivalent harmonic impedance obtained in step (A) into the equivalent circuit equation to reconstruct the background harmonic voltage data, and sort and segment the background harmonic voltage data to obtain several data segments. The corresponding measurement data and the background harmonic voltage data are processed in the same way. (C) The equivalent harmonic impedance of each data segment is calculated using the random independent vector method, thereby obtaining the harmonic impedance of each data segment. This includes: the harmonic voltage equation is expressed as the harmonic voltage Ux equal to the equivalent harmonic impedance of the k-th data segment multiplied by the magnitude of the current vector plus the background harmonic voltage Ex. After removing the mean, both ends are multiplied by the magnitude of the current vector and the mean is taken again to obtain expression one. The harmonic impedance in each data segment is obtained by combining the statistical independence of Ex and the current vector. (D) The objective function is constructed based on minimizing the difference between the calculated impedance value of each data segment obtained in step (C) and the initial impedance value. The optimal initial value that satisfies the objective function is used as the harmonic impedance estimate of the feeder of concern. Finally, the contribution of the harmonic source is quantified by using the estimated equivalent harmonic impedance of the feeder of concern and the vector projection of the harmonic current on the harmonic voltage of the bus of concern.

2. The multi-harmonic source responsibility quantification method based on reconstructed data processing according to claim 1, characterized in that, Step (A) specifically includes the following steps: (A1) When multiple feeders are connected to the busbar of the node of interest, analysis is performed using the Norton equivalent circuit with multiple harmonic sources, wherein the system side and each feeder are equivalent to the parallel connection of the corresponding harmonic current source and harmonic impedance, letting and These are the harmonic impedances of the system side and each feeder, respectively. and Let be the harmonic currents emitted by the harmonic sources on the system side and each feeder, respectively; be the harmonic current of the feeder; and by the superposition theorem, focus on the harmonic voltage at the bus. It is the sum of the harmonic voltages generated at the PCC when each feeder harmonic source acts individually, i.e. Wherein is the harmonic voltage generated at PCC by the harmonic source on the k-th feeder; It is the harmonic voltage generated at the PCC by the system-side harmonic source, i.e., the background harmonic voltage, denoted as... The equivalent impedance excluding system-side harmonic impedance; The equivalent harmonic impedance excluding the harmonic source of feeder k, i.e., the parallel connection of the harmonic impedances of other branches besides feeder k: Harmonic contribution index of feeder k It can be determined by the vector projection of the harmonic voltage it generates at the PCC point onto the total harmonic voltage at the PCC point: The determination of harmonic contribution depends on harmonic impedance. An accurate estimate; (A2) in The target harmonic source feeder i is separated, and the remaining feeder harmonic sources are combined with the background harmonic voltage term. Assuming the harmonic impedance remains constant, the following equation is obtained for the j-th sampled data: in For random error, Consider it as a linear regression model: in ， ， The regression coefficients θ are fitted using the least squares method: Since there is no other known information besides the harmonic measurement data, the initial values ​​of the harmonic impedance are first fitted using the least squares method. 。 3. The multi-harmonic source responsibility quantification method based on reconstructed data processing according to claim 2, characterized in that, The step (B) specifically includes the following steps: (B1) in Divide both ends by phase angle Where α is and The phase angle between them; α0 is and Phase angle between; separation The real and imaginary parts can be obtained as follows: in and Let be the real and imaginary parts of the harmonic impedance, respectively. ， ， ， ,get: (B2) Substitute the initial value of the harmonic impedance obtained in (A1) into The background harmonic voltage data E was reconstructed. x(y) ; background harmonic data E x(y) Sort and segment the data from smallest to largest, and the corresponding measurement data U x(y) , Also according to E x(y) Arrange them in order and divide them into segments.

4. The multi-harmonic source responsibility quantification method based on reconstructed data processing according to claim 3, characterized in that, Step (C) specifically includes the following steps: (C1) In the calculation of the real part, assuming the total number of samples is N, the number of segments is P, and the sample size of each sub-data segment is n=N / P, the harmonic voltage equation for each data segment can be obtained: The subscript k represents the equation of the k-th data segment, E x It is the initial impedance value fitted in step (A2). Substitution The inverse solution, obtained through piecewise division, is U. x and The samples need to be in accordance with E x Arrange the data in the same order and divide them into segments; (C2) The formula for calculating the harmonic impedance in each data segment is: , and the harmonic impedance of each data segment is calculated according to this formula.

5. The multi-harmonic source responsibility quantification method based on reconstructed data processing according to claim 4, characterized in that, The derivation of the formula for calculating the harmonic impedance in each data segment is as follows: The average value is: Multiply both ends simultaneously Then, taking the average, we get: E x This includes background harmonics emitted from the system side and harmonics emitted from other feeder harmonic sources besides feeder i, as well as the statistical independence between different harmonic sources, E. x and If they are independent or only weakly correlated, their covariance is 0, i.e. Will Substitution The formula for calculating the harmonic impedance in each data segment can be obtained as follows: 。 6. The multi-harmonic source responsibility quantification method based on reconstructed data processing according to claim 5, characterized in that, Step (D) specifically includes the following steps: (D1) Define the difference function ΔZ between the calculated impedance and the initial impedance value for each data segment, and its calculation formula is as follows: P represents the number of sub-segments, each input impedance initial value. This will result in a difference function ΔZ. x ΔZ x It is the initial value of impedance. The function; when ΔZ x The minimum value indicates that the difference between the initial impedance value and the calculated impedance value of each data segment is the smallest. This initial impedance value is the optimal initial impedance value, and the objective function is: (D2) Define the impedance change step size δ to iterate the initial impedance value, and determine the direction of iteration by calculating γ. The specific direction of the initial impedance iteration is as follows: (D3) Repeat steps (B1) to (D2) until the difference function ΔZ is obtained. x No further increases were made, and ΔZ was verified within a range of 10 times the initial impedance value. x If it is the minimum value, then the initial impedance value corresponding to the minimum value of the difference function is what the algorithm seeks. If it is not the minimum value, the initial value of the iterative impedance is recalculated. After calculating the harmonic impedance, the harmonic contribution of the harmonic source of feeder i is calculated by the formula.

Images