A distribution network line loss anomaly location method and system considering topological anomalies

By using multiple linear regression models in the distribution network to identify and eliminate abnormal nodes, the problem of low linear loss monitoring accuracy is solved, and the precise positioning of line loss anomalies and the operation efficiency of the power system is improved.

CN118761013BActive Publication Date: 2024-12-03STATE GRID ZHEJIANG ELECTRIC POWER CO LTD HANGZHOU POWER SUPPLY CO +2
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411237665.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-05
Publication Date
2024-12-03
Estimated Expiration
2044-09-05

AI Technical Summary

Technical Problem

The prior art has low accuracy in line loss monitoring and positioning in distribution networks, and it is impossible to effectively identify various types of line loss abnormalities. It lacks real-time and initiative, making it difficult to adapt to the needs of fast response and refined management of modern smart grids.

Method used

Using multiple linear regression models based on node measurement data, multiple linear regression models are constructed to identify and eliminate topological abnormal nodes and natural line loss abnormal nodes, and perform line loss abnormal positioning step by step through iterative analysis to improve the accuracy of line loss calculation .

Benefits of technology

It has achieved accurate positioning of abnormal wire loss in distribution networks, improved the operating efficiency and maintenance level of the power system, and provided strong support for the safe and economic operation of the power system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118761013B_ABST
    Figure CN118761013B_ABST
Patent Text Reader

Abstract

The present invention discloses a method and system for locating abnormal line loss of a distribution network considering topological anomalies, the method comprising: calculating the overall line loss of the distribution network based on the node measurement data of the distribution network, the overall line loss including the natural line loss of the distribution network under normal topological conditions and the management line loss caused by the abnormal topological conditions of the distribution network; constructing a first linear regression model according to the management line loss; constructing a second linear regression model based on the node withdrawal rule, and identifying and eliminating all topological abnormal nodes based on the first linear regression model and the second linear regression model; constructing a third linear regression model according to the natural line loss; constructing a fourth linear regression model based on the node withdrawal rule, and identifying and eliminating all natural line loss abnormal nodes based on the third linear regression model and the fourth linear regression model. The present invention realizes efficient and accurate positioning of the abnormal line loss of the distribution network by comprehensively considering the abnormal natural line loss and the abnormal management line loss, thereby improving the operating efficiency and maintenance level of the power system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent distribution network, and in particular to a method and system for locating abnormal line loss in a distribution network taking into account topological abnormalities. Background Art

[0002] With the rapid development of smart grids, the distribution network, as an important part of the power system, has a direct impact on the quality and stability of power supply due to its operating efficiency and maintenance level. Line loss management is one of the key indicators used to measure the economy and operating status of the power grid. It represents the power loss from the power generation end to the user end, including both technical and management losses. Abnormal line loss usually refers to line loss beyond the normal range, which may be caused by equipment failure, line aging, illegal access, etc. Effective monitoring and positioning of line loss abnormalities are of great significance for preventing power theft, identifying abnormal grid topology, and improving the overall operational efficiency of the grid.

[0003] Traditional line loss management methods mainly rely on manual regular inspections and simple statistical analysis. These methods have problems such as slow response, inaccurate positioning, and difficulty in discovering hidden line loss anomalies. In recent years, although some studies have introduced automated monitoring equipment and preliminary data analysis technologies, the processing and analysis capabilities of large amounts of data in a complex and changeable distribution network environment are still limited, especially in terms of quickly identifying abnormal patterns and accurately distinguishing normal losses from abnormal losses. Moreover, existing line loss management methods usually focus on post-analysis, lack real-time and initiative, and are difficult to adapt to the rapid response and refined management needs of modern smart grids. In particular, in the face of the new situation of large-scale distributed energy access and intensified changes in user-side loads, how to quickly and accurately identify and locate line loss anomalies in massive data has become a key challenge to improving the efficiency and intelligence level of power system operation and maintenance.

[0004] Therefore, there is an urgent need for a distribution network line loss abnormality monitoring and analysis technology that can adapt to the needs of modern smart grids. Summary of the invention

[0005] The purpose of the present invention is to overcome the limitations of the prior art that the line loss monitoring and positioning accuracy is not high and various types of line loss anomalies cannot be effectively identified. In order to achieve the above purpose, the present invention provides a distribution network line loss anomaly positioning method and system considering topological anomalies.

[0006] In a first aspect, an embodiment of the present invention provides a method for locating abnormal line loss in a distribution network considering topological abnormalities, including:

[0007] Based on the node measurement data of the distribution network, the overall line loss of the distribution network is calculated, where the overall line loss includes the natural line loss of the distribution network under normal topology and the management line loss caused by abnormal topology of the distribution network;

[0008] Constructing a first linear regression model according to the management line loss, wherein the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node;

[0009] A second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and eliminated based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first subnode to the last subnode in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each subnode after the node is withdrawn, and the current of each subnode after the node is withdrawn;

[0010] Constructing a third linear regression model according to the natural line loss, wherein the third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network;

[0011] A fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line losses are identified and eliminated based on the third linear regression model and the fourth linear regression model. The fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn.

[0012] Preferably, after constructing the fourth linear regression model based on the node withdrawal rule, and identifying and eliminating all nodes with abnormal natural line loss based on the third linear regression model and the fourth linear regression model, the method further includes:

[0013] Based on the steps of N iterative positioning, the topological abnormal nodes and natural line loss abnormal nodes of each level of the distribution network are obtained; wherein N is the number of levels of the distribution network;

[0014] Each iteration of positioning steps includes:

[0015] According to the current distribution network level, updating the headend node and the subnodes of the distribution network;

[0016] Based on the node measurement data of the updated distribution network, the overall line loss of the updated distribution network is calculated, wherein the overall line loss includes the natural line loss of the updated distribution network under the condition of normal topology and the management line loss caused by the abnormal topology of the updated distribution network;

[0017] Constructing a first linear regression model according to the management line loss, wherein the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node;

[0018] A second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and eliminated based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first subnode to the last subnode in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each subnode after the node is withdrawn, and the current of each subnode after the node is withdrawn;

[0019] Constructing a third linear regression model according to the natural line loss, wherein the third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network;

[0020] A fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line losses are identified and eliminated based on the third linear regression model and the fourth linear regression model. The fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn.

[0021] Preferably, the calculating the overall line loss of the distribution network based on the node measurement data of the distribution network includes:

[0022] Obtain node measurement data of the distribution network;

[0023] Calculating the natural line loss using an equivalent impedance method based on the node measurement data;

[0024] According to the node measurement data, the management line loss is calculated using the following formula; wherein the formula is:

[0025]

[0026] in, Indicates management line loss. Represents the power sequence of the head-end node at different time intervals, represents the power sequence of child node j at different time intervals, represents the power sequence of child node i at different time intervals, Indicates the number of topological nodes, Indicates the number of abnormal nodes in the topology.

[0027] Preferably, constructing a first linear regression model according to the management line loss includes:

[0028] Linear regression is used to perform data fitting on the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node to obtain a first linear regression model; wherein the first linear regression model is characterized by the following formula:

[0029]

[0030] in, Indicates management line loss. represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the power sequence of different time intervals of child node j, represents the power sequence of child node j at different time intervals, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, represents the error term.

[0031] Preferably, the second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and removed based on the first linear regression model and the second linear regression model, including:

[0032] Based on n iterations, all topological abnormal nodes are identified and removed; where n represents the number of topological nodes;

[0033] The steps of each iteration include:

[0034] Based on the node withdrawal rule, update the withdrawal node; wherein the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence;

[0035] Linear regression is used to perform data fitting on the management line loss after the node is withdrawn, the head node current of the distribution network, the power of each sub-node after the node is withdrawn, and the current of each sub-node after the node is withdrawn to obtain a second linear regression model;

[0036] According to the first linear regression model, the power coefficient and the current coefficient corresponding to the withdrawn node are calculated by the least square method; wherein the power coefficient is the coefficient of the power sequence of the withdrawn node at different time intervals, and the current coefficient is the coefficient of the current sequence of the withdrawn node at different time intervals;

[0037] Using the least square method to calculate the first coefficient of determination of the first linear regression model and the second coefficient of determination of the second linear regression model respectively;

[0038] According to the power coefficient, the current coefficient, the first determination coefficient and the second determination coefficient, it is determined whether the withdrawn node is a topological abnormal node, and if so, the withdrawn node is eliminated.

[0039] Preferably, judging whether the withdrawn node is a topologically abnormal node according to the power coefficient, the current coefficient, the first determination coefficient and the second determination coefficient includes:

[0040] According to the power coefficient and the current coefficient, judging whether the withdrawn node meets the first topology abnormal node determination condition; wherein the first topology abnormal node determination condition is that both the power coefficient and the current coefficient are less than a first threshold value;

[0041] According to the first determination coefficient and the second determination coefficient, judging whether the withdrawn node meets the second topological abnormal node determination condition; wherein the second topological abnormal node determination condition is that the difference between the second determination coefficient and the first determination coefficient is higher than a second threshold value;

[0042] It is determined whether the withdrawn node satisfies both the first topology abnormal node determination condition and the second topology abnormal node determination condition. If so, the withdrawn node is determined to be a topology abnormal node.

[0043] Preferably, constructing a third linear regression model according to the natural line loss includes:

[0044] Linear regression is used to perform data fitting on the natural line loss and the head-end node current and each sub-node current of the distribution network to obtain a third linear regression model; wherein the third linear regression model is characterized by the following formula:

[0045]

[0046] in, represents the natural line loss, represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, represents the error term.

[0047] Preferably, the fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line loss are identified and eliminated based on the third linear regression model and the fourth linear regression model, including:

[0048] Based on n iterations, all nodes with abnormal natural line loss are identified and eliminated; where n represents the number of topological nodes;

[0049] The steps of each iteration include:

[0050] Based on the node withdrawal rule, update the withdrawal node; wherein the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence;

[0051] Using linear regression to perform data fitting on the natural line loss after the node is withdrawn, the head node current of the distribution network, and the current of each sub-node after the node is withdrawn, to obtain a fourth linear regression model;

[0052] According to the third linear regression model, the current coefficient corresponding to the withdrawn node is calculated by using the least square method; wherein the current coefficient is the coefficient of the current sequence of the withdrawn node at different time intervals;

[0053] Using the least square method to calculate the third coefficient of determination of the third linear regression model and the fourth coefficient of determination of the fourth linear regression model respectively;

[0054] According to the current coefficient, the third determination coefficient and the fourth determination coefficient, it is determined whether the withdrawal node is a node with abnormal natural line loss, and if so, the withdrawal node is eliminated.

[0055] Preferably, judging whether the withdrawal node is a node with abnormal natural line loss according to the current coefficient, the third determination coefficient and the fourth determination coefficient includes:

[0056] According to the current coefficient, judging whether the withdrawn node meets the first natural line loss abnormal node determination condition; wherein the first natural line loss abnormal node determination condition is that the current coefficient is greater than a third threshold;

[0057] According to the third determination coefficient and the fourth determination coefficient, judging whether the withdrawal node meets the second natural line loss abnormal node determination condition; wherein the second natural line loss abnormal node determination condition is that the difference between the third determination coefficient and the fourth determination coefficient is higher than a fourth threshold value;

[0058] It is determined whether the evacuation node satisfies both the first natural line loss abnormal node determination condition and the second natural line loss abnormal node determination condition. If so, the evacuation node is determined to be a natural line loss abnormal node.

[0059] In a second aspect, an embodiment of the present invention provides a distribution network line loss anomaly location system considering topology anomalies, including:

[0060] An overall line loss calculation module, used to calculate the overall line loss of the distribution network based on the node measurement data of the distribution network, wherein the overall line loss includes the natural line loss of the distribution network under normal topology and the management line loss caused by abnormal distribution network topology;

[0061] A first linear regression model building module, used to build a first linear regression model according to the management line loss, wherein the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node;

[0062] A topology abnormal node identification and elimination module is used to construct a second linear regression model based on the node withdrawal rule, and identify and eliminate all topology abnormal nodes based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first subnode to the last subnode in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each subnode after the node is withdrawn, and the current of each subnode after the node is withdrawn;

[0063] A third linear regression model construction module, used to construct a third linear regression model according to the natural line loss, wherein the third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network;

[0064] The natural line loss abnormal node identification and elimination module is used to construct a fourth linear regression model based on the node withdrawal rule, and identify and eliminate all natural line loss abnormal nodes based on the third linear regression model and the fourth linear regression model. The fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn.

[0065] Compared with the prior art, the method and system for locating abnormal line loss in a distribution network considering topological anomalies in an embodiment of the present invention have the following beneficial effects: a multivariate linear regression model is used to identify abnormal topological nodes and abnormal natural line loss nodes in the distribution network, so that the line loss anomaly in the distribution network can be effectively located; at the same time, eliminating these abnormal nodes can effectively eliminate the influence of the abnormal nodes on the line loss calculation, thereby improving the accuracy of the line loss calculation; through iterative analysis, the line loss anomaly is located step by step, so that the line loss anomaly in the distribution network can be accurately located, thereby improving the operation efficiency and maintenance level of the power system, and providing strong support for the safe and economical operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a flow chart of a method for locating abnormal line loss in a distribution network considering topological abnormality according to an embodiment of the present invention;

[0067] Figure 2 is a schematic diagram of a process for calculating overall line loss according to an embodiment of the present invention;

[0068] Figure 3 is a schematic diagram of a node structure of a distribution network according to an embodiment of the present invention;

[0069] Figure 4 It is a schematic diagram of a process of identifying and removing abnormal topological nodes according to an embodiment of the present invention;

[0070] Figure 5 It is a schematic diagram of a process for determining abnormal topology nodes according to an embodiment of the present invention;

[0071] Figure 6 It is a schematic diagram of a process of identifying and eliminating abnormal nodes of natural line loss according to an embodiment of the present invention;

[0072] Figure 7 It is a schematic diagram of a process for determining a node with abnormal natural line loss according to an embodiment of the present invention;

[0073] Figure 8 It is another flow chart of a method for locating abnormal line loss in a distribution network considering topology abnormality according to an embodiment of the present invention;

[0074] Fig. 9 is a schematic diagram of iteratively locating abnormal nodes according to an embodiment of the present invention;

[0075] Fig.10 It is a structural schematic diagram of a distribution network line loss anomaly locating system considering topology anomalies according to an embodiment of the present invention;

[0076] Fig. 9 In the figure, A represents the main feeder node; a, b, c represent branch line nodes; a1, a2, b1, b2, b3, c1, c2, c3 represent distribution transformer nodes. DETAILED DESCRIPTION

[0077] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0078] like Figure 1 As shown, an embodiment of the present invention provides a method for locating abnormal line loss in a distribution network considering topological abnormalities, comprising the steps of:

[0079] S1. Calculate the overall line loss of the distribution network based on the node measurement data of the distribution network;

[0080] Specifically, Figure 2 As shown, step S1 includes:

[0081] S101, obtaining node measurement data of the distribution network;

[0082] A node refers to a device or user with line loss in a distribution network, including a head-end node and a sub-node. The head-end node and the sub-node can be a main feeder node and a branch line node, a branch line node and a distribution transformer node, or a distribution transformer node and a user node. For a detailed diagram of the node structure of a distribution network, see Figure 3 Specifically, the measurement data of the node is collected by a smart meter, wherein the measurement data at least includes power and current.

[0083] S102. Calculate natural line loss using the equivalent impedance method based on the node measurement data;

[0084] From a statistical point of view, the overall line loss is equal to the power of the head-end node minus the sum of the power of all the end nodes, that is:

[0085]

[0086] in, Represents the overall line loss, Represents the power sequence of the head-end node at different time intervals, represents the power sequence of child node j at different time intervals, Indicates the number of topological nodes. It can be understood that a topological node is a child node, that is, an end node.

[0087] The overall line loss includes the natural line loss of the distribution network under normal topology and the management line loss caused by abnormal distribution network topology. Among them, when the management line loss is not taken into account and the distribution network topology is normal, the equivalent impedance method is used to calculate the natural line loss. Specifically, the following formula is used to calculate the natural line loss:

[0088]

[0089] in, represents the natural line loss, represents the natural line loss of child node i, represents the impedance of child node i, It represents the square of the current flowing through the child node i. When the line loss of the child node i is abnormal, it will be expressed as impedance abnormal.

[0090] S103. Calculate management line loss based on node measurement data.

[0091] This step mainly considers the management line loss caused by abnormal distribution network topology. The abnormal distribution network topology refers to the connection error between the sub-node and the head-end node. When there is a node topology abnormality, the management line loss is calculated using the following formula:

[0092]

[0093] in, Indicates management line loss. Represents the power sequence of the head-end node at different time intervals, represents the power sequence of child node j at different time intervals, represents the power sequence of child node i at different time intervals, Indicates the number of topological nodes, Indicates the number of abnormal nodes in the topology.

[0094] S2. constructing a first linear regression model based on the management line loss;

[0095] Specifically, linear regression is used to fit the data of the management line loss and the head node current of the distribution network, the power of each sub-node, and the current of each sub-node to obtain the first linear regression model. Among them, the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head node current of the distribution network, the power of each sub-node, and the current of each sub-node. Specifically, the following formula is used to characterize the first linear regression model:

[0096]

[0097] in, Indicates management line loss. represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the power sequence of different time intervals of child node j, represents the power sequence of child node j at different time intervals, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, Represents the error term, which represents an unobservable random variable. In this embodiment, It follows a normal distribution with mean 0.

[0098] S3, constructing a second linear regression model based on the node withdrawal rule, and identifying and removing all topological abnormal nodes based on the first linear regression model and the second linear regression model;

[0099] Specifically, based on n iterations, all topological abnormal nodes are identified and removed, where n represents the number of topological nodes.

[0100] Each iteration step, such as Figure 4 As shown, including:

[0101] S301, based on the node withdrawal rule, update the withdrawal node;

[0102] Specifically, the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence. According to the node withdrawal rule, the withdrawal node of the current iteration step is selected.

[0103] S302, using linear regression to perform data fitting on the management line loss after the node is withdrawn, the head node current of the distribution network, the power of each sub-node after the node is withdrawn, and the current of each sub-node after the node is withdrawn, to obtain a second linear regression model;

[0104] The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each sub-node after the node is withdrawn, and the current of each sub-node after the node is withdrawn. Specifically, the following formula is used to characterize the second linear regression model:

[0105]

[0106] in, represents the management line loss after withdrawing node k, represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the power sequence of different time intervals of child node j, represents the power sequence of child node j at different time intervals, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, Represents the error term, which represents an unobservable random variable. In this embodiment, It follows a normal distribution with mean 0.

[0107] S303, according to the first linear regression model, using the least square method to calculate the power coefficient and current coefficient corresponding to the withdrawn node;

[0108] Specifically, the power coefficient is a coefficient of a power sequence at different time intervals of the withdrawn node, and the current coefficient is a coefficient of a current sequence at different time intervals of the withdrawn node.

[0109] S304, using the least square method to calculate the first determination coefficient of the first linear regression model and the second determination coefficient of the second linear regression model respectively;

[0110] The coefficient of determination is used to measure the quality of model fit, and its value ranges from 0 to 1. The larger the coefficient of determination, the stronger the model's ability to explain the dependent variable, that is, the better the model fits.

[0111] S305: judging whether the withdrawn node is a topologically abnormal node according to the power coefficient, the current coefficient, the first determination coefficient and the second determination coefficient, and if so, removing the withdrawn node.

[0112] When the withdrawn node is a topologically abnormal node, its power and current have a large negative contribution to the management line loss, the power coefficient and current coefficient of the withdrawn node are close to -1, and the increase in the determination coefficient after the node is withdrawn is significantly increased.

[0113] Specifically, Figure 5 As shown, step S305 includes:

[0114] S305-A, judging whether the withdrawn node meets the first topology abnormal node determination condition according to the power coefficient and the current coefficient;

[0115] The first topology abnormal node determination condition is that both the power coefficient and the current coefficient are less than the first threshold. The first threshold set in this embodiment is -0.8, that is, the first topology abnormal node determination condition is that both the power coefficient and the current coefficient are less than -0.8.

[0116] S305-B, judging whether the withdrawn node satisfies the second topology abnormal node determination condition according to the first determination coefficient and the second determination coefficient;

[0117] The second topology abnormal node determination condition is that the difference between the second determination coefficient and the first determination coefficient is greater than the second threshold. The second threshold set in this embodiment is 0.4, that is, the second topology abnormal node determination condition is that the difference between the second determination coefficient and the first determination coefficient is greater than 0.4.

[0118] S305-C, determining whether the withdrawn node satisfies both the first topology abnormal node determination condition and the second topology abnormal node determination condition. If so, determining that the withdrawn node is a topology abnormal node.

[0119] If the withdrawn node satisfies both the first topology abnormal node determination condition and the second topology abnormal node determination condition, the withdrawn node is determined to be a topology abnormal node.

[0120] S4, constructing the third linear regression model based on natural line loss;

[0121] Linear regression is used to fit the data of natural line loss and the first-end node current of the distribution network and each sub-node current to obtain a third linear regression model. The third linear regression model is used to characterize the corresponding relationship between the natural line loss and the first-end node current of the distribution network and each sub-node current. Specifically, the following formula is used to characterize the third linear regression model:

[0122]

[0123] in, represents the natural line loss, represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, Represents the error term, which represents an unobservable random variable. In this embodiment, It follows a normal distribution with mean 0.

[0124] S5. Construct a fourth linear regression model based on the node withdrawal rule, and identify and eliminate all nodes with abnormal natural line losses based on the third linear regression model and the fourth linear regression model.

[0125] Specifically, based on n iterations, all nodes with abnormal natural line losses are identified and eliminated, where n represents the number of topological nodes.

[0126] Each iteration step, such as Figure 6 As shown, including:

[0127] S501, based on the node withdrawal rule, update the withdrawal node;

[0128] Specifically, the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence. According to the node withdrawal rule, the withdrawal node of the current iteration step is selected.

[0129] S502, using linear regression to perform data fitting on the natural line loss after the node is withdrawn, the head node current of the distribution network, and the current of each sub-node after the node is withdrawn, to obtain a fourth linear regression model;

[0130] The fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn. Specifically, the following formula is used to characterize the fourth linear regression model:

[0131]

[0132] in, represents the natural line loss after withdrawing node k, represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, Represents the error term, which represents an unobservable random variable. In this embodiment, It follows a normal distribution with mean 0.

[0133] S503, calculating the current coefficient corresponding to the withdrawal node using the least square method according to the third linear regression model;

[0134] Specifically, the current coefficient is a coefficient of a current sequence of different time intervals of the withdrawn node.

[0135] S504, respectively calculating the third coefficient of determination of the third linear regression model and the fourth coefficient of determination of the fourth linear regression model by using the least square method;

[0136] The coefficient of determination is used to measure the quality of model fit, and its value ranges from 0 to 1. The larger the coefficient of determination, the stronger the model's ability to explain the dependent variable, that is, the better the model fits.

[0137] S505. Determine whether the withdrawn node is a node with abnormal natural line loss according to the current coefficient, the third determination coefficient and the fourth determination coefficient. If so, remove the withdrawn node.

[0138] When the withdrawn node is a node with abnormal natural line loss, its current has a large positive contribution to the natural line loss, the current coefficient of the withdrawn node is close to 1, and the decrease in the determination coefficient after the node is withdrawn increases significantly.

[0139] Specifically, Figure 7 As shown, step S505 includes:

[0140] S505-A, judging whether the withdrawn node meets the first natural line loss abnormal node determination condition according to the current coefficient;

[0141] The first natural line loss abnormal node determination condition is that the current coefficient is greater than the third threshold. The third threshold set in this embodiment is 0.7, that is, the first natural line loss abnormal node determination condition is that the current coefficient is greater than 0.7.

[0142] S505-B, judging whether the withdrawn node satisfies the second natural line loss abnormal node determination condition according to the third determination coefficient and the fourth determination coefficient;

[0143] The second natural line loss abnormal node determination condition is that the difference between the third determination coefficient and the fourth determination coefficient is greater than a fourth threshold. The fourth threshold set in this embodiment is 0.5, that is, the second natural line loss abnormal node determination condition is that the difference between the third determination coefficient and the fourth determination coefficient is greater than 0.5.

[0144] S505-C, determine whether the withdrawn node satisfies both the first natural line loss abnormal node determination condition and the second natural line loss abnormal node determination condition. If so, determine that the withdrawn node is a natural line loss abnormal node.

[0145] If the evacuated node satisfies both the first natural line loss abnormal node determination condition and the second natural line loss abnormal node determination condition, the evacuated node is determined to be a natural line loss abnormal node.

[0146] In order to further accurately locate abnormal topological nodes and abnormal natural line loss nodes, such as Figure 8 As shown, after step S5, the following steps are also included:

[0147] S6. Based on the steps of N iterative positioning, the topological abnormal nodes and natural line loss abnormal nodes at each level of the distribution network are obtained.

[0148] Where N is the number of levels of the distribution network.

[0149] Specifically, each iterative positioning step includes:

[0150] 1) Update the headend nodes and subnodes of the distribution network according to the current distribution network level;

[0151] 2) Based on the node measurement data of the updated distribution network, the overall line loss of the updated distribution network is calculated. The overall line loss includes the natural line loss of the updated distribution network under normal topology and the management line loss caused by abnormal topology of the updated distribution network;

[0152] 3) A first linear regression model is constructed according to the management line loss. The first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node;

[0153] 4) A second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and eliminated based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first sub-node to the last sub-node in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each sub-node after the node is withdrawn, and the current of each sub-node after the node is withdrawn;

[0154] 5) A third linear regression model is constructed based on the natural line loss. The third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network;

[0155] 6) A fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line loss are identified and eliminated based on the third linear regression model and the fourth linear regression model. The fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head node current of the distribution network and the current of each sub-node after the node is withdrawn.

[0156] It should be noted that step S6 adopts a top-down step-by-step analysis method, starting from the main feeder node and gradually refining to the branch line node and distribution transformer node. Through iterative positioning, the topological abnormal nodes and natural line loss abnormal nodes at each level of the distribution network are determined, so as to accurately locate the specific location of the line loss abnormality. The iterative positioning process can be seen in Fig. 9 .

[0157] Furthermore, in order to continuously improve the line loss abnormality positioning accuracy and operation and maintenance efficiency of the distribution network, after step S6, the method further includes:

[0158] 1) Feedback the abnormal node identification results to the distribution network operation and maintenance personnel, guide them to conduct on-site inspections, verify the accuracy of the identification results, and repair or adjust the confirmed abnormal nodes;

[0159] 2) Update database information based on on-site verification results, and use feedback data to optimize the model to improve the accuracy and efficiency of future anomaly positioning.

[0160] The embodiment of the present invention provides a method for locating abnormal line loss in a distribution network taking into account topological anomalies. A multivariate linear regression model is used to identify abnormal topological nodes and abnormal natural line loss nodes in the distribution network, and the line loss anomaly in the distribution network can be effectively located. At the same time, eliminating these abnormal nodes can effectively eliminate the influence of the abnormal nodes on the line loss calculation, thereby improving the accuracy of the line loss calculation. Through iterative analysis, the line loss anomaly is located step by step, and the accurate positioning of the line loss anomaly in the distribution network can be achieved, thereby improving the operating efficiency and maintenance level of the power system, and providing strong support for the safe and economical operation of the power system.

[0161] Based on the above distribution network line loss abnormality location method considering topological anomalies, Fig.10 As shown, an embodiment of the present invention provides a distribution network line loss abnormality location system considering topology abnormality, including:

[0162] The overall line loss calculation module 1 is used to calculate the overall line loss of the distribution network based on the node measurement data of the distribution network. The overall line loss includes the natural line loss of the distribution network under normal topology and the management line loss caused by abnormal distribution network topology;

[0163] A first linear regression model construction module 2, used to construct a first linear regression model according to the management line loss, the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node;

[0164] The topological abnormal node identification and elimination module 3 is used to construct a second linear regression model based on the node withdrawal rule, and identify and eliminate all topological abnormal nodes based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first sub-node to the last sub-node in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each sub-node after the node is withdrawn, and the current of each sub-node after the node is withdrawn;

[0165] A third linear regression model construction module 4 is used to construct a third linear regression model according to natural line loss, and the third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network;

[0166] The natural line loss abnormal node identification and elimination module 5 is used to construct a fourth linear regression model based on the node withdrawal rule, and identify and eliminate all natural line loss abnormal nodes based on the third linear regression model and the fourth linear regression model. The fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn.

[0167] It should be noted that each module in the above-mentioned distribution network line loss anomaly location system considering topological anomalies can be fully or partially implemented by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules. For the specific definition of a distribution network line loss anomaly location system considering topological anomalies, please refer to the above definition of a distribution network line loss anomaly location method considering topological anomalies. The two have the same functions and effects, which will not be repeated here.

[0168] In summary, the embodiment of the present invention provides a method and system for locating line loss anomalies in a distribution network taking into account topological anomalies. A multivariate linear regression model is used to identify topological abnormal nodes and natural line loss abnormal nodes in the distribution network, and the line loss anomalies in the distribution network can be effectively located. At the same time, eliminating these abnormal nodes can effectively eliminate the influence of abnormal nodes on line loss calculation, thereby improving the accuracy of line loss calculation. Through iterative analysis, line loss anomalies are located step by step, which can achieve accurate positioning of distribution network line loss anomalies, improve the operating efficiency and maintenance level of the power system, and provide strong support for the safe and economical operation of the power system.

[0169] Each embodiment in this specification is described in a progressive manner, and the same or similar parts of each embodiment can be directly referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. It should be noted that the technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above-mentioned embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0170] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and substitutions can be made without departing from the technical principles of the present invention. These improvements and substitutions should also be regarded as the scope of protection of the present invention.

Claims

1. A method for locating abnormal line loss in a distribution network considering topological anomalies, characterized in that: include: Based on the node measurement data of the distribution network, the overall line loss of the distribution network is calculated, where the overall line loss includes the natural line loss of the distribution network under normal topology and the management line loss caused by abnormal topology of the distribution network; Constructing a first linear regression model according to the management line loss, wherein the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node; A second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and eliminated based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first subnode to the last subnode in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each subnode after the node is withdrawn, and the current of each subnode after the node is withdrawn; Constructing a third linear regression model according to the natural line loss, wherein the third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network; A fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line losses are identified and eliminated based on the third linear regression model and the fourth linear regression model, wherein the fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn; The constructing a first linear regression model according to the management line loss includes: Linear regression is used to perform data fitting on the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node to obtain a first linear regression model; wherein the first linear regression model is characterized by the following formula: in, Indicates management line loss. represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the power sequence of different time intervals of child node j, represents the power sequence of child node j at different time intervals, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, represents the error term; The second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and removed based on the first linear regression model and the second linear regression model, including: Based on n iterations, all topological abnormal nodes are identified and removed; where n represents the number of topological nodes; The steps of each iteration include: Based on the node withdrawal rule, update the withdrawal node; wherein the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence; Linear regression is used to perform data fitting on the management line loss after the node is withdrawn, the head node current of the distribution network, the power of each sub-node after the node is withdrawn, and the current of each sub-node after the node is withdrawn to obtain a second linear regression model; According to the first linear regression model, the power coefficient and the current coefficient corresponding to the withdrawn node are calculated by the least square method; wherein the power coefficient is the coefficient of the power sequence of the withdrawn node at different time intervals, and the current coefficient is the coefficient of the current sequence of the withdrawn node at different time intervals; Using the least square method to calculate the first coefficient of determination of the first linear regression model and the second coefficient of determination of the second linear regression model respectively; According to the power coefficient, the current coefficient, the first determination coefficient and the second determination coefficient, determining whether the withdrawn node is a topological abnormal node, and if so, eliminating the withdrawn node; The step of judging whether the withdrawn node is a topologically abnormal node according to the power coefficient, the current coefficient, the first determination coefficient and the second determination coefficient includes: According to the power coefficient and the current coefficient, judging whether the withdrawn node meets the first topology abnormal node determination condition; wherein the first topology abnormal node determination condition is that both the power coefficient and the current coefficient are less than a first threshold value; According to the first determination coefficient and the second determination coefficient, judging whether the withdrawn node meets the second topological abnormal node determination condition; wherein the second topological abnormal node determination condition is that the difference between the second determination coefficient and the first determination coefficient is higher than a second threshold value; Determine whether the withdrawn node satisfies both the first topology abnormal node determination condition and the second topology abnormal node determination condition, and if so, determine that the withdrawn node is a topology abnormal node; The constructing a third linear regression model according to the natural line loss includes: Linear regression is used to perform data fitting on the natural line loss and the head-end node current and each sub-node current of the distribution network to obtain a third linear regression model; wherein the third linear regression model is characterized by the following formula: in, represents the natural line loss, represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, represents the error term; The fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line loss are identified and eliminated based on the third linear regression model and the fourth linear regression model, including: Based on n iterations, all nodes with abnormal natural line loss are identified and eliminated; where n represents the number of topological nodes; The steps of each iteration include: Based on the node withdrawal rule, update the withdrawal node; wherein the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence; Using linear regression to perform data fitting on the natural line loss after the node is withdrawn, the head node current of the distribution network, and the current of each sub-node after the node is withdrawn, to obtain a fourth linear regression model; According to the third linear regression model, the current coefficient corresponding to the withdrawn node is calculated by using the least square method; wherein the current coefficient is the coefficient of the current sequence of the withdrawn node at different time intervals; Using the least square method to calculate the third coefficient of determination of the third linear regression model and the fourth coefficient of determination of the fourth linear regression model respectively; According to the current coefficient, the third determination coefficient and the fourth determination coefficient, determining whether the withdrawal node is a node with abnormal natural line loss, and if so, eliminating the withdrawal node; The step of judging whether the withdrawal node is a node with abnormal natural line loss according to the current coefficient, the third determination coefficient and the fourth determination coefficient includes: According to the current coefficient, judging whether the withdrawn node meets the first natural line loss abnormal node determination condition; wherein the first natural line loss abnormal node determination condition is that the current coefficient is greater than a third threshold; According to the third determination coefficient and the fourth determination coefficient, judging whether the withdrawal node meets the second natural line loss abnormal node determination condition; wherein the second natural line loss abnormal node determination condition is that the difference between the third determination coefficient and the fourth determination coefficient is higher than a fourth threshold value; It is determined whether the evacuation node satisfies both the first natural line loss abnormal node determination condition and the second natural line loss abnormal node determination condition. If so, the evacuation node is determined to be a natural line loss abnormal node.

2. The method for locating abnormal line loss in a distribution network according to claim 1, characterized in that: After constructing the fourth linear regression model based on the node withdrawal rule, and identifying and eliminating all nodes with abnormal natural line loss based on the third linear regression model and the fourth linear regression model, the method further includes: Based on the steps of N iterative positioning, the topological abnormal nodes and natural line loss abnormal nodes of each level of the distribution network are obtained; wherein N is the number of levels of the distribution network; Each iteration of positioning steps includes: According to the current distribution network level, updating the headend node and the subnodes of the distribution network; Based on the node measurement data of the updated distribution network, the overall line loss of the updated distribution network is calculated, wherein the overall line loss includes the natural line loss of the updated distribution network under the condition of normal topology and the management line loss caused by the abnormal topology of the updated distribution network; Constructing a first linear regression model according to the management line loss, wherein the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node; A second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and eliminated based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first subnode to the last subnode in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each subnode after the node is withdrawn, and the current of each subnode after the node is withdrawn; Constructing a third linear regression model according to the natural line loss, wherein the third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network; A fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line losses are identified and eliminated based on the third linear regression model and the fourth linear regression model. The fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn.

3. The method for locating abnormal line loss in a distribution network according to claim 1, characterized in that: The calculating the overall line loss of the distribution network based on the node measurement data of the distribution network includes: Obtain node measurement data of the distribution network; Calculating the natural line loss using an equivalent impedance method based on the node measurement data; According to the node measurement data, the management line loss is calculated using the following formula; wherein the formula is: in, Indicates management line loss. Represents the power sequence of the head-end node at different time intervals, represents the power sequence of child node j at different time intervals, represents the power sequence of child node i at different time intervals, Indicates the number of topological nodes, Indicates the number of abnormal nodes in the topology.

4. A distribution network line loss anomaly location system considering topological anomalies, characterized in that: include: An overall line loss calculation module, used to calculate the overall line loss of the distribution network based on the node measurement data of the distribution network, wherein the overall line loss includes the natural line loss of the distribution network under normal topology and the management line loss caused by abnormal distribution network topology; A first linear regression model building module, used to build a first linear regression model according to the management line loss, wherein the first linear regression model is used to characterize the corresponding relationship between the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node; A topology abnormal node identification and elimination module is used to construct a second linear regression model based on the node withdrawal rule, and identify and eliminate all topology abnormal nodes based on the first linear regression model and the second linear regression model. The node withdrawal rule is to withdraw one node at a time, and the nodes are selected from the first subnode to the last subnode in sequence. The second linear regression model is used to characterize the corresponding relationship between the management line loss after the node is withdrawn and the head-end node current of the distribution network, the power of each subnode after the node is withdrawn, and the current of each subnode after the node is withdrawn; A third linear regression model construction module, used to construct a third linear regression model according to the natural line loss, wherein the third linear regression model is used to characterize the corresponding relationship between the natural line loss and the head-end node current and each sub-node current of the distribution network; A natural line loss abnormal node identification and elimination module is used to construct a fourth linear regression model based on the node withdrawal rule, and identify and eliminate all natural line loss abnormal nodes based on the third linear regression model and the fourth linear regression model, wherein the fourth linear regression model is used to characterize the correspondence between the natural line loss after the node is withdrawn and the head-end node current of the distribution network and the current of each sub-node after the node is withdrawn; The constructing a first linear regression model according to the management line loss includes: Linear regression is used to perform data fitting on the management line loss and the head-end node current of the distribution network, the power of each sub-node, and the current of each sub-node to obtain a first linear regression model; wherein the first linear regression model is characterized by the following formula: in, Indicates management line loss. represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the power sequence of different time intervals of child node j, represents the power sequence of child node j at different time intervals, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, represents the error term; The second linear regression model is constructed based on the node withdrawal rule, and all topological abnormal nodes are identified and removed based on the first linear regression model and the second linear regression model, including: Based on n iterations, all topological abnormal nodes are identified and removed; where n represents the number of topological nodes; The steps of each iteration include: Based on the node withdrawal rule, update the withdrawal node; wherein the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence; Linear regression is used to perform data fitting on the management line loss after the node is withdrawn, the head node current of the distribution network, the power of each sub-node after the node is withdrawn, and the current of each sub-node after the node is withdrawn to obtain a second linear regression model; According to the first linear regression model, the power coefficient and the current coefficient corresponding to the withdrawn node are calculated by the least square method; wherein the power coefficient is the coefficient of the power sequence of the withdrawn node at different time intervals, and the current coefficient is the coefficient of the current sequence of the withdrawn node at different time intervals; Using the least square method to calculate the first coefficient of determination of the first linear regression model and the second coefficient of determination of the second linear regression model respectively; According to the power coefficient, the current coefficient, the first determination coefficient and the second determination coefficient, determining whether the withdrawn node is a topological abnormal node, and if so, eliminating the withdrawn node; The step of judging whether the withdrawn node is a topologically abnormal node according to the power coefficient, the current coefficient, the first determination coefficient and the second determination coefficient includes: According to the power coefficient and the current coefficient, judging whether the withdrawn node meets the first topology abnormal node determination condition; wherein the first topology abnormal node determination condition is that both the power coefficient and the current coefficient are less than a first threshold value; According to the first determination coefficient and the second determination coefficient, judging whether the withdrawn node meets the second topological abnormal node determination condition; wherein the second topological abnormal node determination condition is that the difference between the second determination coefficient and the first determination coefficient is higher than a second threshold value; Determine whether the withdrawn node satisfies both the first topology abnormal node determination condition and the second topology abnormal node determination condition, and if so, determine that the withdrawn node is a topology abnormal node; The constructing a third linear regression model according to the natural line loss includes: Linear regression is used to perform data fitting on the natural line loss and the head-end node current and each sub-node current of the distribution network to obtain a third linear regression model; wherein the third linear regression model is characterized by the following formula: in, represents the natural line loss, represents the intercept term, The coefficient representing the square of the first-end node current, represents the first-end node current, represents the coefficient of the current sequence of child node i at different time intervals, represents the current sequence of child node i at different time intervals, Indicates the number of topological nodes, represents the error term; The fourth linear regression model is constructed based on the node withdrawal rule, and all nodes with abnormal natural line loss are identified and eliminated based on the third linear regression model and the fourth linear regression model, including: Based on n iterations, all nodes with abnormal natural line loss are identified and eliminated; where n represents the number of topological nodes; The steps of each iteration include: Based on the node withdrawal rule, update the withdrawal node; wherein the node withdrawal rule is to withdraw one node each time, and the nodes are selected from the first child node to the last child node in sequence; Using linear regression to perform data fitting on the natural line loss after the node is withdrawn, the head node current of the distribution network, and the current of each sub-node after the node is withdrawn, to obtain a fourth linear regression model; According to the third linear regression model, the current coefficient corresponding to the withdrawn node is calculated by using the least square method; wherein the current coefficient is the coefficient of the current sequence of the withdrawn node at different time intervals; Using the least square method to calculate the third coefficient of determination of the third linear regression model and the fourth coefficient of determination of the fourth linear regression model respectively; According to the current coefficient, the third determination coefficient and the fourth determination coefficient, determining whether the withdrawal node is a node with abnormal natural line loss, and if so, eliminating the withdrawal node; The step of judging whether the withdrawal node is a node with abnormal natural line loss according to the current coefficient, the third determination coefficient and the fourth determination coefficient includes: According to the current coefficient, judging whether the withdrawn node meets the first natural line loss abnormal node determination condition; wherein the first natural line loss abnormal node determination condition is that the current coefficient is greater than a third threshold; According to the third determination coefficient and the fourth determination coefficient, judging whether the withdrawal node meets the second natural line loss abnormal node determination condition; wherein the second natural line loss abnormal node determination condition is that the difference between the third determination coefficient and the fourth determination coefficient is higher than a fourth threshold value; It is determined whether the evacuation node satisfies both the first natural line loss abnormal node determination condition and the second natural line loss abnormal node determination condition. If so, the evacuation node is determined to be a natural line loss abnormal node.

Citation Information

Patent Citations

  • Line loss analysis method and device applied to power distribution network

    CN112688310A

  • District line loss rate abnormity diagnosis method based on data driving

    CN114862139A