Power distribution network power quality comprehensive evaluation method based on complex network theory

Abstract

The invention discloses a power distribution network electric energy quality comprehensive evaluation method based on a complex network theory. The method comprises the following steps: S1, drawing a directed network graph of a power distribution network; s2, acquiring a complex network index of each node in the directed network graph of the power distribution network, and weighting the complex network index by adopting a CRITIC algorithm to calculate a node importance comprehensive decision value of each node in the power distribution network; s3, combining the actual measurement value of the power quality index of each node in the power distribution network with the node importance comprehensive decision value, respectively calculating a subjective weight value and an objective weight value through a G1 method and a CRITIC method, obtaining a comprehensive weight value through subjective and objective comprehensive weight calculation, and carrying out two-stage weight optimization to obtain a final weight value; and S4, according to the final weight value, performing quantitative evaluation on the power quality of each node in the power distribution network by adopting a fuzzy comprehensive evaluation method. The evaluation method can accurately evaluate the electric energy quality of the complex network power distribution network, and is convenient for improving the electric energy quality in a targeted manner.

Description

Technical Field

[0001] The present invention relates to the technical field of power distribution network, and in particular to a comprehensive evaluation method for power quality of a distribution network based on complex network theory. Background Art

[0002] In recent years, with the development of new power systems, new distribution networks will be connected to more and more distributed resources. A large number of distributed new energy sources such as distributed photovoltaics and distributed wind power connected to the new distribution networks will effectively increase the new energy power generation capacity and electricity, and enhance the new energy dominant position in the new power system.

[0003] At present, with the access of distributed power sources and the application of nonlinear loads, the power quality of new distribution networks has become increasingly complex and severe. On the one hand, the increase in power quality-sensitive electrical appliances has led to an increasing demand for power quality on the user side; on the other hand, the access to more distributed new energy sources has further complicated the network structure of the distribution network, and the complex network also has a certain impact on the power quality of the distribution network. Therefore, in order to ensure the power supply quality and improve the user's power experience, it is necessary to improve the power quality of new distribution networks that are connected to a large number of distributed new energy sources.

[0004] However, research has found that traditional power quality evaluation indicators generally only consider narrow power quality indicators, which do not reflect the importance characteristics of each node in a complex distribution network, resulting in the inability to distinguish nodes of different importance in the scoring. Therefore, it is urgent to develop a means of accurately evaluating the power quality of a new distribution network for a complex distribution network, so as to improve the power quality problems of the new distribution network in a targeted manner in the future. Summary of the invention

[0005] The technical problem to be solved by the present invention is to design a comprehensive evaluation method for power quality of distribution network based on complex network theory, so as to solve the problem that the existing traditional power quality evaluation indicators only consider narrow power quality indicators but cannot reflect the importance characteristics of each node in the complex distribution network.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a comprehensive evaluation method for power quality of distribution network based on complex network theory, which comprises the following steps:

[0007] S1: Draw a directed network graph of the distribution network;

[0008] S2: Obtaining complex network indicators of each node in the directed network graph of the distribution network, and weighting the complex network indicators using the CRITIC algorithm to calculate the node importance comprehensive decision value of each node in the distribution network;

[0009] S3: combining the actual measured value of the power quality index of each node in the distribution network with the comprehensive decision value of the node importance, calculating the subjective weight value and the objective weight value thereof by the G1 method and the CRITIC method respectively, and then performing the subjective and objective comprehensive weight calculation to obtain the comprehensive weight value, and performing the two-stage weight optimization on the comprehensive weight value to obtain the final weight value;

[0010] S4: According to the final weight value, a fuzzy comprehensive evaluation method is used to quantitatively evaluate the power quality of each node in the distribution network.

[0011] In the above technical scheme of the present invention, the comprehensive evaluation method of the power quality of the distribution network based on complex network theory can first draw a simplified directed network diagram of the distribution network according to the actual structure and flow direction of the distribution network; then, based on the complex network theory, obtain the complex network indicators of each node in the directed network diagram of the distribution network, such as: node degree centrality, node betweenness centrality, eigenvector centrality and other complex network indicators, and construct the complex network indicator matrix of this distribution network according to the node number; then, based on the CRITIC (Indicator Importance Correlation Method) algorithm, obtain the node importance comprehensive decision value of each node; finally, combine the node importance comprehensive decision value with the measured value of the selected power quality indicator as a data matrix, perform subjective and objective comprehensive weight calculation based on the G1 (Order Relationship Analysis Method) method and the CRITIC method, and then perform two-stage weight optimization to obtain the final weight, and finally quantitatively evaluate the power quality of each node through the fuzzy comprehensive evaluation method.

[0012] Furthermore, in the comprehensive evaluation method for power quality of a distribution network described in the present invention, in the step S1, there are multiple nodes in the directed network graph of the distribution network, and the nodes are: power stations, distributed power sources, substations, converter stations or load equipment.

[0013] In the above technical solution of the present invention, a directed network diagram of the distribution network can be drawn according to the topological structure and flow direction of the distribution network; and when drawing, multiple nodes will be set specifically according to the type of facilities in the distribution network. These nodes are generally power stations, distributed power sources, substations, converter stations, loads and other facilities, and detailed power quality measurement data can be obtained; wherein, in the above-constructed directed network diagram, the connecting lines between the nodes can be understood as the transmission lines between the nodes, and the direction of the connecting lines can be understood as the flow direction of the corresponding transmission lines.

[0014] Furthermore, in the comprehensive evaluation method for power quality of distribution network described in the present invention, in step S2, the complex network indicators include: node degree centrality, node betweenness centrality and eigenvector centrality.

[0015] The study found that each node in the distribution network of a complex network will have different levels of importance due to factors such as its different position in the complex network and the nodes it connects. Therefore, considering the directional flow of power flow in the distribution network, the power quality problem of the node will affect other connected nodes through the power connection line along the flow. In complex network theory, node degree is used to describe the number of nodes directly connected to a node. Therefore, node degree centrality is selected as a complex network indicator.

[0016] In addition, in addition to the nodes directly connected to the node, the location of the node in the distribution network of the complex network also directly reflects its importance. For this reason, the shortest path algorithm and the number of paths passing through a node can be used to reflect the influence of the node in the entire network and the utilization of the node. Therefore, the node betweenness centrality is selected as a complex network indicator.

[0017] In addition, the importance of a node in a distribution network depends not only on its own node degree, but also on the importance of its neighboring nodes. In a distribution network, if the number of edges flowing into a node is greater, its importance is higher, and this node will pass on higher importance to the neighboring nodes connected to it through directed edges. Therefore, eigenvector centrality is selected as a complex network indicator.

[0018] To sum up the above analysis, in the above technical scheme of the present invention, node degree centrality, node betweenness centrality and eigenvector centrality are selected as complex network indicators of each node in the directed network graph of the distribution network. These indicators can be calculated based on the directed network graph of the distribution network, thereby ensuring the practicability of the complex network indicators.

[0019] Further, in the comprehensive evaluation method for power quality of distribution network described in the present invention, in step S2, the node degree centrality, node betweenness centrality and eigenvector centrality of each node in the directed network graph of the distribution network are obtained, specifically:

[0020] The node degree centrality is:

[0021] in, For Node The node degree centrality of The number of directly connected nodes; is the node out-degree, which means the node is the number of links from the starting point; is the node in-degree, which means that the node is the number of links at the end point;

[0022] The node betweenness centrality is:

[0023] Among them, the node The node betweenness centrality is B i ; δ jk For Node To Node The number of paths under the shortest path algorithm; δ jk (i) is δ jk Passing Node The number of paths;

[0024] The eigenvector centrality is:

[0025]

[0026] in, For Node The eigenvector centrality of For Node The eigenvector centrality of For Node The out-degree of D i The directed edge in the directed network graph points to the node is the set of all nodes of , and d is the damping coefficient.

[0027] Furthermore, in the comprehensive evaluation method for power quality of distribution network described in the present invention, in step S2, the CRITIC algorithm is used to weight the complex network indicators to calculate the node importance comprehensive decision value of each node in the distribution network, specifically:

[0028] S21: establishing a complex network indicator matrix based on the values ​​of the complex network indicators of each node in the distribution network;

[0029] S22: performing initial data processing on the complex network indicator matrix to obtain a complex network indicator standardized matrix;

[0030] S23: Calculate the coefficient of variation of the ath indicator

[0031] S24: Calculate the correlation coefficient r between the a'th indicator and the a"th indicator according to the complex network indicator standardization matrix a'a” , then calculate the conflict quantization value A a ;

[0032] S25: According to the coefficient of variation and conflict quantification value A a Calculate the information amount E a ;

[0033] S26: the amount of information E a Perform normalization to obtain the weight of the ath indicator

[0034] S27: According to the calculation of S23-S26 above, the node importance index weight set is θ * , the calculation formula of the node importance decision value of each node is:

[0035]

[0036] in, represents the weight value of the a-th complex network indicator; y ia Represents the standardized data of the ath indicator of the ith node in the standardized matrix of complex network indicators.

[0037] Furthermore, in the comprehensive evaluation method for power quality of a distribution network described in the present invention, in step S3, the power quality indicators include: voltage deviation, voltage fluctuation, voltage sag, three-phase imbalance, harmonics and frequency deviation.

[0038] Further, in the comprehensive evaluation method for power quality of distribution network of the present invention, in step S3, the actual measured values ​​of the six power quality indicators selected by each node in the distribution network are combined with the comprehensive decision value of the node importance, and a power quality data matrix X is established based on the above seven evaluation indicators. m×l :

[0039]

[0040] Among them, i = (1, 2, 3, ..., m), b = (1, 2, 3, ..., 7), m is the number of nodes in the distribution network, l is the number of evaluation indicators, and l is 7.

[0041] Furthermore, in the method for comprehensive evaluation of power quality of distribution network of the present invention, in step S3:

[0042] The subjective weight is calculated based on the G1 method, specifically: obtaining multiple evaluation indicators including the power quality indicator and the node importance comprehensive decision value, first determining the evaluation indicator sequence relationship, and then determining the relative importance ratio r between the evaluation indicators k , and finally based on the relative importance ratio r among the evaluation indicators k The subjective weight value set of each node is calculated as w;

[0043] Objective weight calculation based on CRITIC method is as follows: m×l Process the data to obtain the standardized matrix of node power quality data; calculate the coefficient of variation of the index; calculate the correlation coefficient between the b'th evaluation index and the b"th evaluation index according to the standardized matrix to calculate the conflict quantification value A b ; According to the conflict quantification value Ab The information volume is calculated by the coefficient of variation of the indicator; the information volume is normalized to obtain the objective weight value θ of the bth indicator b .

[0044] The calculation formula of the comprehensive weight is as follows: The calculation formula of the subjective and objective comprehensive weight λ is:

[0045]

[0046] Among them, λ is the comprehensive weight, w1~w l is the subjective weight of the first to the first indicator calculated based on the G1 method, θ b is the objective weight of the bth indicator calculated based on the CRITIC method.

[0047] Further, in the method for comprehensive evaluation of power quality of distribution network described in the present invention, in step S3, the comprehensive weight value is subjected to two-stage weight optimization to obtain the final weight value, specifically:

[0048] Taking the comprehensive weight value λ as the initial weight, the expected weight value ω of the bth evaluation indicator of the i-th node is calculated b for:

[0049]

[0050] According to the purpose of the two-stage method, the optimization model of the evaluation index weights is established using the least squares method:

[0051]

[0052] ω b ≥0

[0053] Based on the above optimization model, we can obtain W = [ω1, ω2, …, ω l ], which is the final weight value of each evaluation indicator.

[0054] Furthermore, in the comprehensive evaluation method for power quality of distribution network described in the present invention, in step S4, the evaluation index is first quantified to obtain a comment set for quantitative grading of power quality, and then the membership degree of each node data corresponding to the evaluation set is calculated by the membership function, and finally a weighted average fuzzy comprehensive operator is used to perform fuzzy product operation on the evaluation matrix, and the evaluation result is quantified and calculated.

[0055] The beneficial effects of the present invention are as follows: the inventor has optimized and designed a comprehensive evaluation method for the power quality of a distribution network based on complex network theory. Compared with the power quality indicators selected in the traditional method, a complex network indicator is also added, namely, the comprehensive decision value of the node importance of each node in the distribution network, and the differentiated scoring of nodes with different importance can be reflected in the quantitative evaluation score; at the same time, since a two-stage weight optimization is adopted on the basis of the one-stage subjective and objective comprehensive weighting, the rationality of the final weight value obtained can be improved, and the mutual influence of the subjective and objective weights in the one-stage weighting process can be avoided. The subjective and objective weight values ​​can be more reasonably integrated, and a scoring reference basis can be provided for the subsequent quantitative evaluation of the power quality of the distribution network, so as to facilitate the subsequent improvement of the power supply quality of the distribution network and improve the user's power experience. It has good promotion prospects and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The present invention is a flowchart of the steps of the method for comprehensive evaluation of power quality of distribution network in one implementation manner.

[0057] Figure 2 The schematic diagram shows the architecture of a distribution network in one implementation of the method for comprehensive evaluation of power quality in a distribution network according to the present invention.

[0058] Figure 3 A directed network diagram of a distribution network in one implementation manner of the method for comprehensive evaluation of power quality of a distribution network according to the present invention is schematically shown.

[0059] Figure 4 The power quality evaluation index set of the distribution network power quality comprehensive evaluation method according to the present invention in one implementation manner is schematically shown.

[0060] Figure 5 The schematic diagram shows the application of the comprehensive evaluation method of power quality of distribution network of the present invention to Figure 3 The node importance comprehensive decision values ​​of 23 nodes finally calculated in the directed network graph of the distribution network shown.

[0061] Figure 6 The schematic diagram shows the comparison between the power quality comprehensive evaluation method of the distribution network of the present invention and the traditional power quality evaluation method. Figure 3 Evaluation results of 23 nodes in the directed network graph of the distribution network shown. DETAILED DESCRIPTION

[0062] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.

[0063] At present, with the trend of increasingly complex new power system networks, traditional evaluation indicators only consider conventional power quality indicators, but do not reflect the importance characteristics of each node in the complex distribution network, which is not conducive to the subsequent improvement of the power quality of the distribution network. To this end, the present invention proposes a comprehensive evaluation method for distribution network power quality based on complex network theory.

[0064] See also Figure 1 It can be seen that in Figure 1 In the illustrated implementation manner, the comprehensive evaluation method for power quality of a distribution network designed by the present invention specifically comprises the following steps:

[0065] S1: Draw a directed network diagram of the distribution network based on the topological structure and flow direction of the distribution network.

[0066] S2: Obtain the complex network index of each node in the directed network graph of the distribution network, and use the CRITIC algorithm to weight the complex network index to calculate the node importance comprehensive decision value of each node in the distribution network.

[0067] S3: Combine the actual measured value of the power quality index of each node in the distribution network with the comprehensive decision value of the node importance, calculate the subjective weight value and the objective weight value respectively by the G1 method and the CRITIC method, then perform the subjective and objective comprehensive weight calculation to obtain the comprehensive weight value, and perform the two-stage weight optimization on the comprehensive weight value to obtain the final weight value.

[0068] S4: According to the final weight value, a fuzzy comprehensive evaluation method is used to quantitatively evaluate the power quality of each node in the distribution network.

[0069] It is not difficult to see from the above content that, in addition to the selected power quality indicators, the comprehensive evaluation method for the power quality of the distribution network designed by the present invention also adds complex network indicators, namely, the comprehensive decision value of the node importance of each node in the distribution network, and the differentiated scoring of nodes with different importance can be reflected in the quantitative evaluation score; at the same time, since the two-stage weight optimization is adopted on the basis of the one-stage subjective and objective comprehensive weighting, the rationality of the final weight value obtained can be improved, and the mutual influence of the subjective and objective weights in the one-stage weighting process can be avoided. The subjective and objective weight values ​​can be more reasonably integrated, which can provide a scoring reference for the subsequent quantitative evaluation of the power quality of the distribution network.

[0070] like Figure 3 As shown, in the present invention, the directed network diagram of the above distribution network has multiple nodes, which are generally power stations, distributed power sources, substations, converter stations and load equipment on the distribution network.

[0071] See also Figure 2 and Figure 3As shown, in step S1 of the above-mentioned comprehensive evaluation method for power quality of distribution network, when drawing a directed network diagram of the distribution network according to the topological structure and flow direction of the distribution network, the nodes in the directed network diagram can be set specifically according to the types of facilities in the distribution network; wherein, Figure 3 The connecting lines between the nodes shown are the transmission lines between the two facilities. Figure 3 The direction between the nodes shown is the direction of power flow on the transmission line between the two corresponding facilities.

[0072] Accordingly, in the present invention, when executing the above step S2, it is necessary to first calculate the complex network indicator value of each node in the directed network diagram of the distribution network to obtain the complex network indicator matrix of all nodes in this network; then, through a weighting algorithm, consider the correlation between each complex network indicator and assign weights to each complex network indicator; finally, calculate the node importance comprehensive decision value of each node based on the complex network indicator value of each node and the weight value of each complex network indicator.

[0073] Among them, in the present invention, node degree centrality, node betweenness centrality and eigenvector centrality can be specifically selected as complex network indicators of each node in the directed network graph of the distribution network to reflect the nature, function and correlation relationship of each node in the distribution network. These indicators can all be calculated based on the directed network graph of the distribution network, ensuring the practicability of complex network indicators.

[0074] In the present invention, the following calculation method is specifically used to obtain the above three complex network indicators:

[0075] (1) Node degree centrality: The node The node degree centrality is denoted as Its representation is similar to the node The number of directly connected nodes;

[0076]

[0077] in, is the node out-degree, which means the node is the number of links from the starting point; is the node in-degree, which means that the node The number of links to the endpoint.

[0078] It should be pointed out that the above-mentioned “link” is the connecting line in the directed network diagram of the distribution network, that is, Represents a node As the starting point, and with this node The number of connected lines; similarly, Represents a node As the end point, and with this node The number of connected wires.

[0079] (2) Node betweenness centrality: Node betweenness centrality can accurately characterize the influence of a node in the network and the utilization of the node. The node betweenness centrality is B i ;

[0080]

[0081] Among them, the node The node betweenness centrality is B i ; δ jk For Node To Node The number of paths under the shortest path algorithm; δ jk (i) is δ jk Passing Node The number of paths.

[0082] (3) Eigenvector centrality: Based on the PageRank algorithm, it can calculate the nodes in the directed network graph. The eigenvector centrality of Eigenvector centrality The calculation formula is:

[0083]

[0084] in, For Node The eigenvector centrality of For Node The eigenvector centrality of For Node The out-degree of D i The directed edge in the directed network graph points to the node The set of all nodes of , d is the damping coefficient, and the damping coefficient d is generally 0.85.

[0085] Accordingly, see further Figure 1 It can be seen that based on the complex network index values ​​of each node, in the present invention, it is also necessary to use the CRITIC algorithm to weight the complex network index to calculate the node importance comprehensive decision value of each node in the distribution network. Among them, the use of the CRITIC method for node importance evaluation can reflect its objectivity. The CRITIC method can eliminate the influence between indicators with strong correlation, reduce the information overlap between indicators, and obtain a more scientific and reliable evaluation result.

[0086] In practical application, the following steps S21-S27 may be specifically included:

[0087] S21: Based on the values ​​of multiple complex network indicators of each node in the distribution network, a complex network indicator matrix of all nodes in the distribution network is obtained:

[0088] Assume that m power quality monitoring points constitute the sample set S = {S i}(i=1,2,…,m), and n complex network index sets G={g a}(a=1,2,…,n), then the complex network index matrix Ψ of all nodes in the distribution network m×n for:

[0089]

[0090] in, is the value of the ath complex network index of the ith node.

[0091] S22: For the above complex network indicator matrix Ψ m×n After initial data processing, the standardized matrix of complex network indicators Y = (y ia ) m×n :

[0092]

[0093] Where: ia is the a-th complex network index value of the i-th node object, i=1,…,m,a=1,2,3,m is the total number of nodes, maxψ a ,minψ a are the maximum and minimum values ​​of the feature data of different nodes under the ath indicator. ia It is the standardized data value after processing the a-th complex network indicator value of the i-th node.

[0094] S23: Calculating the coefficient of variation of complex network indicators The coefficient of variation is a quantitative display of the comparative strength of complex network indicators. The calculation formula is:

[0095]

[0096] Among them, s a is the standard deviation of the a-th complex network index, is the average value of the a-th complex network index.

[0097] S24: According to the complex network index standardization matrix Y = (y ia ) m×n The correlation coefficient r between the a'th complex network index and the a"th complex network index is calculated. a'a” , then calculate the conflict quantization value Aa :

[0098]

[0099] in, They are the standardized matrices Y=(y ia ) m×n The covariance between the a'th column and the a"th column.

[0100] S25: Based on the coefficient of variation and conflict quantification value A a Calculate the information amount E a :

[0101]

[0102] S26: the amount of information E a Perform normalization to obtain the weight of the ath indicator

[0103]

[0104] S27: According to the calculation of S23-S26 above, the node importance index weight set is θ * , the calculation formula of the node importance decision value of each node is:

[0105]

[0106] Among them, θ a represents the weight value of the a-th complex network indicator; y ia Represents the standardized data of the ath indicator of the ith node in the standardized matrix of complex network indicators.

[0107] Accordingly, see further Figure 1 It can be seen that in step S3 of the present invention, it is necessary to combine the actual measured value of the power quality index of each node in the distribution network with the node importance comprehensive decision value calculated in the above step S27 to establish a power quality data matrix.

[0108] For example, in some embodiments, the power quality indicators can be specifically selected as: voltage deviation, voltage fluctuation, voltage sag, three-phase imbalance, harmonics and frequency deviation, and a total of 7 power quality evaluation indicators are combined with the node importance comprehensive decision value to form the power quality data matrix X m×l :

[0109]

[0110] Among them, i = (1, 2, 3, ..., m), b = (1, 2, 3, ..., 7), m is the number of nodes in the distribution network, l is the number of evaluation indicators, and l is 7.

[0111] At the same time, if Figure 1 As shown, in step S3, a power quality data matrix X is constructed. m×l Finally, the G1 method and CRITIC method can be used for subjective and objective weighting, and the comprehensive weight value can be obtained by performing subjective and objective comprehensive weighting according to the obtained subjective weight and objective weight.

[0112] The G1 method is a method that determines the weight of each evaluation indicator based on the subjective experience of researchers and the importance of each evaluation indicator. Researchers can rank and quantify the importance of indicators based on the characteristics of the active distribution network and knowledge and experience. Among them, the G1 method is used for subjective weighting to obtain the subjective weight of the evaluation indicators, including:

[0113] (1) Determine the order relationship of evaluation indicators: Suppose the evaluation factor set T is composed of l evaluation indicators, which is {T1, T2, …, T l}, if the evaluation index T l' is more important than T l” , then record it as T l' >T l” , rearrange the evaluation indicators from large to small in importance, and determine the order relationship of the evaluation indicators as Where T l * Represents the lth evaluation indicator after sorting by importance.

[0114] (2) Determine the relative importance ratio between evaluation indicators: Determine the adjacent evaluation indicators and Importance R k :

[0115]

[0116] in: and Represents the evaluation indicators and The weight coefficient, r k The value of is: k =1.0 means evaluation index and The importance of r k =1.2 indicates the evaluation index and Slightly important, k =1.4 indicates the evaluation index and Obviously important,k =1.6 indicates the evaluation index and Strongly important, k =1.8 represents the evaluation index and Extremely important, and k When 1.1, 1.3, 1.5, and 1.7 are taken, they represent the middle values ​​of the above adjacent judgments.

[0117] (3) Calculate the subjective weight of the evaluation indicators: According to the relative importance ratio r between the determined evaluation indicators k Use the following formula to calculate the subjective weight value of the lth evaluation indicator:

[0118]

[0119] The subjective weight value of the k-1th evaluation indicator is:

[0120]

[0121] The CRITIC method is used to weight each evaluation indicator in the data matrix. The CRITIC method can eliminate the influence of evaluation indicators with strong correlation, reduce the information overlap between evaluation indicators, and obtain more scientific and reliable evaluation results; the process of using CRITIC for objective weighting and obtaining the objective weight value of the evaluation indicator includes:

[0122] (1) Data standardization: For the power quality data matrix X m×l After data processing, the node power quality data standardization matrix Y = (y ib ) m×l , the specific data processing method is as follows:

[0123]

[0124] Among them, in the power quality data matrix X m×l x in ib is the bth complex network index value of the i-th node object, i=1,…,m, b=1,2,…,7, m is the number of nodes, max(x b ), min(x b ) are the maximum and minimum values ​​of different node feature data under the same complex network index, y ib It is the standardized data value after processing of the bth power quality index value of the ith evaluation object.

[0125] (2) Calculate the coefficient of variation of the evaluation index: The coefficient of variation is a quantitative display of the comparative strength of the index. The coefficient of variation Vσ b The calculation formula is:

[0126]

[0127] Among them, s b is the standard deviation of the b-th evaluation indicator, is the average value of the bth evaluation indicator.

[0128] (3) Calculation of conflict of evaluation indicators: According to the standardized matrix Y = (y ib ) m×l Calculate the correlation coefficient r between the b'th evaluation index and the b"th evaluation index b'b” , then calculate the conflict quantization value A b :

[0129]

[0130]

[0131] in: is the covariance between the b'th column and the b"th column of the standardized matrix.

[0132] (4) Calculation of information volume: The information volume E is calculated using the following formula: b :

[0133] E b =Vσ b ×A b

[0134] (5) Calculation of objective weights of evaluation indicators: Normalize the amount of information to obtain the weight θ of the bth evaluation indicator b for:

[0135]

[0136] In order to avoid the problem of over-reliance on expert opinions caused by a single subjective weighting method and over-reliance on quantitative analysis of sample data and neglect of subjective qualitative analysis caused by a single objective weighting method, a combined weighting method combining subjective weighting method (G1 method) and objective weighting method (CRITIC method) is used to calculate the comprehensive weight λ. The calculation formula of the subjective and objective comprehensive weight value λ is:

[0137]

[0138] Among them, λ is the comprehensive weight, w1~w l is the subjective weight of the first to the first indicator calculated based on the G1 method, θ b is the objective weight of the bth indicator calculated based on the CRITIC method.

[0139] According to the above-mentioned first stage, the corresponding comprehensive weight value set of evaluation indicators is obtained through subjective weighting method, objective weighting method and subjective and objective comprehensive weighting method. Finally, the two-stage weighting method is used to optimize the above comprehensive weight values ​​in the second stage to calculate the final weight value.

[0140] For example: Taking the comprehensive weight value λ obtained by the subjective and objective comprehensive weighting method as the initial weight, the expected weight value of the bth evaluation indicator of the i-th node is calculated for:

[0141]

[0142] Evaluation object node S i The comprehensive decision value f i The weighted calculation is

[0143] According to the purpose of the two-stage method, suppose there is a weight ω of the final evaluation index b (b=1,2,…,l) can make the comprehensive decision value of all samples under the final weight Combined decision value f of all nodes i sum With the smallest difference, the optimization model of the evaluation index weights is established using the least squares method:

[0144]

[0145] ω b ≥0

[0146] Based on the above optimization model, we can obtain W = [ω1, ω2, …, ω l ], which is the final weight value of each evaluation indicator.

[0147] It should be noted that in the present invention, when performing the optimization solution, the corresponding code is written in Matlab based on the optimization model objective function and constraints, and the result is obtained by directly running the program. Therefore, there is no need to elaborate on the calculation of the optimization model here.

[0148] Accordingly, see further Figure 1 It can be seen that in step S4 of the present invention, it is necessary to first quantify the power quality evaluation index to obtain a comment set for quantitative grading of power quality, and then calculate the membership degree of each node data corresponding to the judgment set through the membership function, and finally use the weighted average fuzzy comprehensive operator to perform fuzzy product operation on the final weight value and the judgment matrix, and quantify the evaluation result.

[0149] S41. Quantification of power quality evaluation indicators:

[0150] The fuzzy comprehensive evaluation first determines the factor set and the comment set. The factor set T composed of the indicators confirmed in step S3 is: {T1 (voltage deviation), T2 (voltage fluctuation), T3 (voltage sag), T4 (three-phase imbalance), T5 (harmonics), T6 (frequency deviation), T7 (node ​​importance comprehensive decision value)};

[0151] The power quality is divided into five different levels, forming a five-level fuzzy evaluation set V: {V1 (excellent power quality), V2 (good power quality), V3 (average power quality), V4 (poor power quality), V5 (poor power quality)}.

[0152] Through the fuzzy judgment set quantitative evaluation index, the corresponding relationship between the quantitative grading of the power quality of the distribution network and the scoring interval is obtained as follows: Level V1 indicates excellent power quality, the corresponding scoring interval is (85,100], and its quantitative score is 95; Level V2 indicates good power quality, the corresponding scoring interval is (75,85], and its quantitative score is 80; Level V3 indicates average power quality, the corresponding scoring interval is (60,75], and its quantitative score is 70; Level V4 indicates poor power quality, the corresponding scoring interval is (50,60], and its quantitative score is 55; Level V5 indicates poor power quality, the corresponding scoring interval is (0,50], and its quantitative score is 40.

[0153] The resulting review set V is:

[0154] V=[Excellent, Good, Average, Poor, Bad]

[0155] =[95 80 70 55 40].

[0156] S42. Quantification of power quality evaluation indicators:

[0157] The membership of the data in the power quality evaluation index matrix corresponding to different comments can be calculated by the membership function, using the Gauss membership function. The Gauss membership function f(y) is expressed as follows:

[0158]

[0159] Where: y ib is the standardized data of the jth power quality evaluation index of the ith node, σ and c are the two parameters of the Gauss membership function. In this paper, σ is taken as 0.3, and the value of c represents the center position of the membership function. Five c values ​​are taken: c1 = 1, c2 = 0.75, c3 = 0.5, c4 = 0.25, c5 = 0 to ensure that each indicator has 5 comment memberships.

[0160] Substitute the parameters σ and c into the membership function to obtain the membership calculation formula corresponding to the five evaluation sets. ib Substitute them into the membership functions of the five evaluation levels respectively, and the evaluation matrix J is obtained as follows:

[0161]

[0162] in: It represents the indicator y ib For the judging level V t degree of affiliation.

[0163] S43, fuzzy comprehensive evaluation of power quality:

[0164] use The operator (weighted average fuzzy comprehensive operator) performs fuzzy product operation on the weight and the evaluation matrix to obtain the overall evaluation of the new distribution network power quality evaluation system as follows:

[0165] B i =[b i (V1)b i (V2)b i (V3)b i (V4)b i (V5)]

[0166] Among them, B i represents the membership set of the five comments corresponding to the i-th node object; b i (V t ) represents the power quality index of each node i relative to the tth comment V t The membership degree is calculated as where ω b is the final weight of the bth indicator

[0167] Finally, the evaluation result is quantified and the quantized score Z of the i-th node is calculated. i The calculation formula is:

[0168]

[0169] According to the calculated power quality evaluation result and the correspondence between the quantitative classification and scoring interval of the power quality grade of the above distribution network, a comment on the power quality of the distribution network is obtained.

[0170] Embodiment 1:

[0171] In order to verify the superiority and reliability of the comprehensive evaluation method for power quality of distribution network designed by the present invention, the inventors also specifically tested and verified and listed the actual evaluation results of Example 1 for verification.

[0172] This embodiment is based on Figure 1-Figure 4 The distribution network, directed network diagram and evaluation method shown in FIG. 1 are implemented. The directed network diagram of the distribution network of the first embodiment is as follows: Figure 3 As shown, the directed network diagram of the distribution network has a total of 23 nodes, each of which can monitor the data of the power quality indicators. The data of the monitored power quality indicators are shown in the following Table 1.

[0173] Table 1 lists the data of six power quality indicators, namely, voltage deviation, voltage fluctuation, voltage sag, three-phase imbalance, harmonics and frequency deviation, of 23 nodes in the directed network diagram of the distribution network of Example 1.

[0174] Table 1.

[0175]

[0176]

[0177] Based on the contents in Table 1 above, the node degree centrality, betweenness centrality and eigenvector centrality of each node are calculated using the scheme of step S2 of the comprehensive evaluation method for power quality of distribution network of the present invention, and the node importance comprehensive decision value of each node is calculated based on the CRITIC weighting method. The node importance comprehensive decision values ​​of the 23 nodes finally obtained are as follows: Figure 5 As shown in the bar graph.

[0178] According to the calculated node importance comprehensive decision values ​​of the 23 nodes and the actual measured values ​​of the six power quality indicators of voltage deviation, voltage fluctuation, voltage sag, three-phase unbalance, harmonics and frequency deviation of the 23 nodes listed in Table 1 above, the subjective weight value and objective weight value are calculated respectively by the G1 method and the CRITIC method, and then the subjective and objective comprehensive weight calculation is performed to obtain the comprehensive weight value.

[0179] Then the two-stage weighting method is used to assign weights to obtain the final weight value, and then the fuzzy comprehensive evaluation method is used to quantitatively evaluate the power quality of each node in the distribution network to obtain the final quantitative evaluation result.

[0180] For the convenience of comparison, the present invention also lists a quantitative evaluation structure for evaluating the distribution network of the first embodiment using a traditional power quality evaluation method without considering the importance characteristics of each node in the distribution network of a complex network.

[0181] Table 2 lists the evaluation scores of the 23 nodes of the distribution network using the comprehensive evaluation method for power quality of the distribution network of the present invention (taking into account the importance of each node in the distribution network) and the traditional power quality evaluation method (not taking into account the importance of each node in the distribution network) in the evaluation level corresponding to each node under the division of the comment set V.

[0182] Table 2.

[0183]

[0184]

[0185] Among them, in the comment set V, level V1 indicates excellent power quality, the corresponding scoring interval is (85,100], and its quantitative score is 95; level V2 indicates good power quality, the corresponding scoring interval is (75,85], and its quantitative score is 80; level V3 indicates average power quality, the corresponding scoring interval is (60,75], and its quantitative score is 70; level V4 indicates poor power quality, the corresponding scoring interval is (50,60], and its quantitative score is 55; level V5 indicates poor power quality, the corresponding scoring interval is (0,50], and its quantitative score is 40.

[0186] Accordingly, Figure 6 The comparison between the evaluation results of the distribution network power quality comprehensive evaluation method of the present invention taking into account the node importance and the evaluation results of the traditional power quality evaluation means not taking into account the comprehensive decision value of the node importance is shown; wherein, referring to nodes 6, 8, and 15, the scoring results of the distribution network power quality comprehensive evaluation method of the present invention are lower than those of the traditional power quality evaluation means not taking into account the node importance, while the scores of nodes with lower node importance, such as nodes 7, 10, and 13, are not significantly reduced or improved, thus achieving the purpose of the present invention of differentiated evaluation of the power quality of nodes with higher importance.

[0187] At the same time, it can be seen from Table 2 that the evaluation results of the comprehensive evaluation method for power quality of distribution network considering node importance adopted by the present invention are basically consistent with the evaluation method not considering node importance, and have high adaptability. For node 6 with high node importance, due to the constraint of its high node importance comprehensive decision value, the rating is reduced by one level in the fuzzy comprehensive evaluation process, reaching a stricter evaluation standard.

[0188] In summary, the comprehensive evaluation method for power quality of distribution network based on complex network theory of the present invention can solve the problem that the existing traditional power quality evaluation indicators only consider the narrow power quality indicators, but cannot reflect the importance characteristics of each node in the complex distribution network. When the comprehensive evaluation method for power quality of distribution network is actually applied, it not only considers the traditional power quality indicators, but also adds the indicator of complex network parameters, so that the differentiated scores of nodes with different importance can be reflected in the quantitative evaluation score. Therefore, the comprehensive evaluation method for power quality of distribution network can accurately evaluate the power quality of the distribution network of the complex network, so as to improve the power supply quality of the distribution network in the future and improve the power consumption experience of the user. It has good promotion prospects and application value.

[0189] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A comprehensive evaluation method for power quality of distribution network based on complex network theory, characterized in that: The steps include: S1: Draw a directed network graph of the distribution network; S2: Obtaining complex network indicators of each node in the directed network graph of the distribution network, and weighting the complex network indicators using the CRITIC algorithm to calculate the node importance comprehensive decision value of each node in the distribution network; S3: combining the actual measured value of the power quality index of each node in the distribution network with the comprehensive decision value of the node importance, calculating the subjective weight value and the objective weight value thereof by the G1 method and the CRITIC method respectively, and then performing the subjective and objective comprehensive weight calculation to obtain the comprehensive weight value, and performing the two-stage weight optimization on the comprehensive weight value to obtain the final weight value; S4: According to the final weight value, a fuzzy comprehensive evaluation method is used to quantitatively evaluate the power quality of each node in the distribution network.

2. The method for comprehensive evaluation of power quality of distribution network according to claim 1, characterized in that: In the step S1, there are a plurality of nodes in the directed network graph of the distribution network, and the nodes are: power stations, distributed power sources, substations, converter stations or load equipment.

3. The method for comprehensive evaluation of power quality of distribution network according to claim 1, characterized in that: In step S2, the complex network indicators include: node degree centrality, node betweenness centrality and eigenvector centrality.

4. The method for comprehensive evaluation of power quality of distribution network according to claim 3, characterized in that: In step S2, the node degree centrality, node betweenness centrality and eigenvector centrality of each node in the directed network graph of the distribution network are obtained, specifically: The node degree centrality is: in, For Node The node degree centrality of The number of directly connected nodes; is the node out-degree, which means the node is the number of links from the starting point; is the node in-degree, which means that the node is the number of links at the end point; The node betweenness centrality is: Among them, the node The node betweenness centrality is B i ; δ jk For Node To Node The number of paths under the shortest path algorithm; δ jk (i) is δ jk Passing Node The number of paths; The eigenvector centrality is: in, For Node The eigenvector centrality of For Node The eigenvector centrality of For Node The out-degree of D i The directed edge in the directed network graph points to the node is the set of all nodes of , and d is the damping coefficient.

5. The method for comprehensive evaluation of power quality of distribution network according to claim 4, characterized in that: In step S2, the CRITIC algorithm is used to weight the complex network indicators to calculate the node importance comprehensive decision value of each node in the distribution network, specifically: S21: establishing a complex network indicator matrix based on the values ​​of the complex network indicators of each node in the distribution network; S22: performing initial data processing on the complex network indicator matrix to obtain a complex network indicator standardized matrix; S23: Calculate the coefficient of variation of the ath indicator S24: Calculate the correlation coefficient r between the a'th indicator and the a"th indicator according to the complex network indicator standardization matrix a'a” , then calculate the conflict quantization value A a ; S25: According to the coefficient of variation and conflict quantification value A a Calculate the information amount E a ; S26: the amount of information E a Perform normalization to obtain the weight of the ath indicator S27: According to the calculation of S23-S26 above, the node importance index weight set is θ * , the calculation formula of the node importance decision value of each node is: in, represents the weight value of the a-th complex network indicator; y ia Represents the standardized data of the ath indicator of the ith node in the standardized matrix of complex network indicators.

6. The method for comprehensive evaluation of power quality of distribution network according to claim 1, characterized in that: In step S3, the power quality indicators include: voltage deviation, voltage fluctuation, voltage sag, three-phase imbalance, harmonics and frequency deviation.

7. The method for comprehensive evaluation of power quality of distribution network according to claim 6, characterized in that: In step S3, the actual measured values ​​of the six power quality indicators selected by each node in the distribution network are combined with the comprehensive decision value of the node importance, and a power quality data matrix X is established based on the above seven evaluation indicators. m×l : Among them, i = (1, 2, 3, ..., m), b = (1, 2, 3, ..., 7), m is the number of nodes in the distribution network, l is the number of evaluation indicators, and l is 7.

8. The method for comprehensive evaluation of power quality of distribution network according to claim 7, characterized in that: In step S3: The subjective weight is calculated based on the G1 method, specifically: obtaining multiple evaluation indicators including the power quality indicator and the node importance comprehensive decision value, first determining the evaluation indicator sequence relationship, and then determining the relative importance ratio r between the evaluation indicators k , and finally based on the relative importance ratio r among the evaluation indicators k The subjective weight value set of each node is calculated as w; Objective weight calculation based on CRITIC method is as follows: m×l Perform data processing to obtain a standardized matrix of node power quality data; Calculate the coefficient of variation of the indicator; calculate the correlation coefficient between the b'th evaluation indicator and the b"th evaluation indicator according to the standardized matrix to calculate the conflict quantitative value A b ; According to the conflict quantification value A b The information volume is calculated by the coefficient of variation of the indicator; the information volume is normalized to obtain the objective weight value θ of the bth indicator b ; The calculation formula of comprehensive weight is as follows: The calculation formula of subjective and objective comprehensive weight λ is: Among them, λ is the comprehensive weight, w1~w l is the subjective weight of the first to the first indicator calculated based on the G1 method, θ b is the objective weight of the bth indicator calculated based on the CRITIC method.

9. The method for comprehensive evaluation of power quality of distribution network according to claim 8, characterized in that: In step S3, the comprehensive weight value is subjected to a two-stage weight optimization to obtain a final weight value, specifically: Taking the comprehensive weight value λ as the initial weight, the expected weight value of the bth evaluation indicator of the i-th node is calculated: for: According to the purpose of the two-stage method, the optimization model of the evaluation index weights is established using the least squares method: oh b ≥0 Based on the above optimization model, we can obtain W = [ω1, ω2, …, ω l ], which is the final weight value of each evaluation indicator.

10. The method for comprehensive evaluation of power quality of distribution network according to claim 7, characterized in that: In step S4, the evaluation index is first quantified to obtain a comment set for quantitative grading of power quality, and then the membership degree of each node data corresponding to the judgment set is calculated through the membership function. Finally, a weighted average fuzzy comprehensive operator is used to perform fuzzy product operation on the final weight value and the judgment matrix, and the evaluation result is quantified.

Images