Low-voltage transformer area topology error correction method and system

Abstract

The application relates to a low-voltage transformer area topology structure error correction method and system. The method comprises the following steps: acquiring a topology relationship matrix of each node in a low-voltage transformer area, wherein the nodes comprise monitoring nodes and user nodes; acquiring power data differences of each node in several rounds, wherein the power data difference is the difference between the power data collected in the current round and the last round; based on the topology relationship matrix, an association relationship formula of the power data differences between the nodes is constructed; the proportion of the times that the power data difference of each monitoring node does not satisfy the association relationship formula is calculated, and is compared with an abnormal frequency threshold value, if the abnormal frequency threshold value is exceeded, the current monitoring node is taken as an abnormal monitoring node; and the topology structure containing the abnormal monitoring node is corrected. The application can effectively meet the low-voltage transformer area topology structure error correction demand, and is easy to implement, and does not need to make a large number of modifications to the original system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of low-voltage transformer area topology structure, in particular to a low-voltage transformer area topology structure error correction method and system. BACKGROUND

[0002] As the last link of power distribution, the low-voltage transformer area has always been difficult to manage due to the large number of users, the variety of loads, the complex line-house relationship, the difficulty in tracing the change of rights and responsibilities, and other reasons. With the rapid development of the economy and society, it has become an urgent problem to establish a clear and reliable topology relationship in low-voltage power supply services.

[0003] For a long time, the relationship between the transformer and the line has been checked and solved by manual investigation. The existing investigation methods include power-off verification and carrier verification. These two methods are tedious and inefficient, and they cannot solve the problem of temporary adjustment of the line. Therefore, many places have begun to use the collected power consumption information data to identify through data calculation. The existing identification methods include voltage identification, current identification, and active power identification. Since it does not require human intervention and hardware modification, it has received more and more attention. However, due to the influence of the accuracy of existing equipment, the computing power of edge computing nodes, and the accuracy of the algorithm itself, it is difficult to balance accuracy and usability. Therefore, it is a technical problem for technicians in this field to develop a practical and effective topology identification algorithm. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a low-voltage transformer area topology structure error correction method and system, which can effectively meet the low-voltage transformer area topology structure error correction requirements and is easy to implement without the need for extensive modification of the original system.

[0005] The technical solution adopted by the present application to solve the technical problem is to provide a low-voltage transformer area topology structure error correction method, comprising the following steps:

[0006] Obtain the topology relationship matrix of each node in the low-voltage transformer area, the nodes including monitoring nodes and user nodes;

[0007] Obtain the power data difference of each node for several rounds, the power data difference being the difference between the power data collected in the current round and the previous round;

[0008] Based on the topology relationship matrix, an association relationship formula of the power data difference between nodes is constructed;

[0009] Calculate the proportion of the number of times that the power data difference of each monitoring node does not satisfy the association relationship formula, and compare it with the abnormal frequency threshold. If it exceeds the abnormal frequency threshold, the current monitoring node is regarded as an abnormal monitoring node;

[0010] Correcting a topology structure containing abnormal monitoring nodes.

[0011] Further, the correcting the topology structure containing abnormal nodes comprises:

[0012] Obtaining user nodes associated with the abnormal monitoring nodes as abnormal user nodes;

[0013] Filtering data associated with the abnormal user nodes from the collected power data differences to obtain a to-be-calculated data set;

[0014] Extracting abnormal user nodes with the largest power data difference variation in a set number of rounds in the to-be-calculated data set as target correction nodes;

[0015] According to the correlation coefficients of the target correction nodes and each abnormal monitoring node, finding the associated monitoring nodes of the target correction nodes and updating the to-be-calculated data set according to the matching results, and then returning to the previous step until all abnormal user nodes are matched to the associated monitoring nodes;

[0016] Updating the topology relationship matrix according to each target correction node and its associated monitoring node.

[0017] Further, the finding the associated monitoring nodes of the target correction nodes according to the correlation coefficients of the target correction nodes and each abnormal user node comprises:

[0018] Calculating the Pearson correlation coefficients of the target correction nodes and each abnormal monitoring node;

[0019] Selecting the two abnormal monitoring nodes with the highest and second-highest Pearson correlation coefficients;

[0020] Calculating the ratio of the highest Pearson correlation coefficient to the second-highest Pearson correlation coefficient, and if the ratio is greater than a set threshold, considering that the abnormal monitoring node corresponding to the highest Pearson correlation coefficient is the associated monitoring node of the target correction node.

[0021] Further, the updating the to-be-calculated data set according to the matching results comprises:

[0022] If there is an associated monitoring node, excluding the data associated with the target correction node from the to-be-calculated data set;

[0023] If there is no associated monitoring node, using the collected power data differences to update the to-be-calculated data set.

[0024] Further, the to-be-calculated data set is represented as

[0025]

[0026] Wherein, L' o is the to-be-calculated data set, L oL is the power data difference of the monitoring node, M' is the topological structure relationship matrix M after excluding the abnormal monitoring node and its associated nodes.

[0027] Further, the target error correction node is represented as

[0028] u=(j|MAX[∑(||l′ uj | / ∑l′ u |)),l′ u ∈L′ u

[0029] Wherein, u is the target error correction node, l′ u is the power data difference of any target error correction node, l′ uj is the power data difference of the jth target error correction node, L′ u is the data set to be calculated.

[0030] Further, the topological relationship matrix is a matrix composed of 1 and 0, with the monitoring node as the column node, and the monitoring node and the user node as the row node, wherein the non-0 element represents that the column node and the row node have an associated relationship.

[0031] Further, the associated relationship formula is represented as

[0032] (L o -ΔL o ) T ≤M{[L o L u ]} T ≤(L o +ΔL o ) T

[0033] Wherein, L o is the power data difference of the monitoring node, ΔL o is the power measurement error of the monitoring node, L u is the power data difference of the user node, ΔL u is the power measurement error of the user node, and M is the topological relationship matrix.

[0034] The application also provides a low-voltage area topology error correction system, characterized by comprising:

[0035] A first input module is used to obtain the topological relationship matrix of each node in the low-voltage area, and the nodes include monitoring nodes and user nodes;

[0036] A second input module is used to obtain the power data difference of each node in several rounds, and the power data difference is the difference between the power data collected in the current round and the last round;

[0037] The correlation module is configured to construct a correlation relationship of the power difference between the nodes based on the topological relationship matrix.

[0038] The screening module is configured to calculate a proportion of times that the power difference of each monitoring node does not satisfy the correlation relationship, and compare the proportion with an abnormal frequency threshold value, and if the proportion exceeds the abnormal frequency threshold value, the current monitoring node is regarded as an abnormal monitoring node.

[0039] The correction function module is configured to correct the topological structure containing the abnormal monitoring node.

[0040] Further, the monitoring nodes include intelligent transformers, intelligent branch boxes and intelligent electrical switches.

[0041] Advantages

[0042] Compared with the prior art, the present application has the following advantages and positive effects: the present application can monitor the topological structure and correct abnormal relationships in the terminal (edge node) of the power supply area by long-term continuous power monitoring data information without modifying the low-voltage hardware equipment, and can determine the accuracy of the original topological structure; compared with the existing clustering grouping scheme based on voltage gradient difference, the present application does not need to improve the device accuracy, and compared with the typical scheme of calculating the power difference by using genetic algorithm or traversal algorithm, the present application reduces the demand for computing power under the premise of ensuring accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is a flowchart of the first embodiment of the present application;

[0044] Figure 2 is a modified topological structure diagram of the first embodiment of the present application. DETAILED DESCRIPTION

[0045] The present application will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present application and not to limit the scope of the present application. In addition, it should be understood that after reading the content taught by the present application, those skilled in the art can make various modifications or changes to the present application, and these equivalent forms also fall within the scope of the claims attached hereto.

[0046] The first embodiment of the present application relates to a topological structure correction method based on Pearson correlation coefficient and data screening, which collects low-voltage user meter information based on Internet of Things technology, monitors the established low-voltage topological structure through the collected power data information of the intelligent fusion terminal of the power supply area, and corrects the abnormal topological structure through correction methods and other technical means, so as to achieve the monitoring and correction effect of the low-voltage power supply area topological structure.

[0047] The specific steps are described as follows.

[0048] S1, the transformer area intelligent fusion terminal starts the monitoring function module based on the existing topology structure, and outputs the detection result.

[0049] The topology structure error correction system is mainly APP, which is installed in the transformer area intelligent fusion terminal, and runs once every period of time, and the detection function is completed through the use of a plurality of rounds of low-voltage feeder monitoring devices, user electric energy meter power consumption data statistical calculation. The transformer, branch box and other intelligent equipment collection points are monitoring nodes, and the user intelligent electric energy meter collection points are user nodes. The existing topology structure is manually initialized and constructed by low-voltage operation and maintenance personnel.

[0050] Firstly, the correlation relationship algorithm formula based on the power consumption change difference is established as follows:

[0051] (L o -ΔL o ) T ≤M{[L o L u ]} T ≤(L o +ΔL o ) T

[0052] In the formula,

[0053] L o —The set of power consumption data difference of the monitoring node between the current round and the last round, L o =[l o1 ,l o2 ,l o3 ,…,l oi ,…], the higher the current round data is positive, and vice versa is negative;

[0054] ΔL o —The set of power consumption measurement error of the monitoring node, which is a percentage parameter determined by the collection equipment error;

[0055] L u —The set of power consumption data difference of the user node between the current round and the last round, L u =[l u1 ,l u2 ,l u3 ,…,l uj ,…], the higher the current round data is positive, and vice versa is negative;

[0056] ΔL u —The set of power consumption measurement error of the user node, which is a percentage parameter determined by the user electric energy meter error;

[0057] M—topology relationship matrix composed of 1 and 0, non-0 element indicates that the column number corresponding node and the row number corresponding node have an association relationship, here the corresponding node represents the monitoring node and the user node arranged in order.

[0058] Then, it is judged whether the data collected by each node meets the association relationship constructed based on the existing topology structure, and for L o set element, representing that the topology relationship related to the node number is abnormal and needs to be output for processing.

[0059] S2, for the abnormal topology relationship, the correction function module is started in time, and the correction is carried out periodically and the corrected topology relationship is output, such as Figure 1 As shown in the figure, the topology correction process includes the following steps:

[0060] Firstly, for the nodes involved in the non-abnormal topology relationship, based on the current data difference set, the abnormal user data part is screened:

[0061]

[0062] In the formula,

[0063] M′—topology relationship matrix M excluding the abnormal part;

[0064] L′ o —abnormal part set of the monitoring node power consumption data difference.

[0065] Secondly, for the abnormal data information, the user data of the specified period is accumulated to form a to-be-calculated data set;

[0066] Thirdly, for the to-be-calculated data set, based on the absolute value, the proportion of each user node difference value in each period is counted, and the user with the largest proportion in the accumulation specified period is screened, that is, the user node with the largest change in the statistical period is screened;

[0067] u=(j|MAX[∑(|l′ uj | / ∑l′ u |)])

[0068] In the formula,

[0069] u—user node number with the largest change in the statistical period;

[0070] l u —user node power consumption data difference value of this round and the last round;

[0071] L′ u —user data set excluding the confirmed correct data.

[0072] Here, ∑|l′u | is the sum of the absolute values of the difference in power consumption of the user nodes in the current period, |l′ uj | / ∑|l′ u | is the sum of the absolute values of the difference in power consumption of the user nodes in the current period, |l′ uj | / ∑|l′ u | is the sum of the absolute values of the difference in power consumption of the user nodes in the current period, |l′ uj | / ∑|l′ u | is the sum of the absolute values of the difference in power consumption of the user nodes in the current period, |l′

[0073] Step 4, based on the data set to be calculated, for the data of the user node smaller than the data of the monitoring node, calculate the Pearson coefficient of the selected user node and the data difference set of the monitoring node, find out the monitoring nodes with the closest and second closest relationship with the user node;

[0074] Step 5, compare the selected monitoring node correlation, for the points with a difference value higher than a certain proportion relationship, determine that the user node and the monitoring node have a correlation relationship, and continue to the next step; otherwise, continue to accumulate the next period data outside the current data set, and re-calculate according to the third to fifth steps to clarify the correlation relationship;

[0075] Step 6, exclude the determined user node and monitoring node in the data set to be calculated, exclude the correct relationship, and return to step 3 for calculation until all user nodes are calculated.

[0076] S3, according to the corrected topological relationship, perform monitoring in the specified period through the monitoring function module, confirm the correctness of the corrected topological relationship and report.

[0077] Next, take a low-voltage area as an example. The actual household relationship of this area is shown in Figure 2 , which covers 4 monitoring nodes and 6 user nodes. The monitoring nodes use low-voltage feeder monitoring devices to ensure relatively accurate and reliable topological structure. The user nodes are equipped with electric energy meters, and the related topological relationship is initialized and constructed by low-voltage operation and maintenance personnel manually, and all are connected to the area intelligent fusion terminal to report data. Data is collected every 15 minutes, and the collection accuracy is 1 level.

[0078] S1, the monitoring function module is started regularly, and the detection result is output.

[0079] Among them, the existing topological relationship matrix M structure formed by low-voltage operation and maintenance personnel manually is as follows:

[0080]

[0081] Among them, the multi-round collection data of this area is as follows, and the measurement error is 1% here:

[0082]

[0083] Wherein, the calculation can be continuously obtained:

[0084]

[0085] According to the abnormal frequency exceeding half as a boundary, it can be determined that the topology related user relationship of o2, o4 two monitoring nodes exists error.

[0086] S2, the correction function module is started, and the positive topology relationship is output. The correction function process is as follows:

[0087] S21, the topology structure relationship matrix M excluding the abnormal part is constructed:

[0088]

[0089] S22, the user data of the specified period is accumulated, here 3 hours 12 rounds are collected, and the abnormal user data part is screened, and the to-be-calculated data set is formed;

[0090]

[0091]

[0092] S23, for the to-be-calculated data set, based on the absolute value, the difference value data proportion of each user node in each period is counted, and all proportion values of the specified period are accumulated, the to-be-processed user 1, 2, 3, 4 data is 2.5195, 3.6446, 2.0453, 3.7905, then the user 4 with the largest change in the screening statistical period is selected;

[0093] S24, the Pearson correlation coefficient of each user can be calculated as 0.04791, 0.99996, -0.73219, 0.000575, and the most closely and the second closely monitored points are node 2 and node 1 respectively;

[0094] S25, the coefficient difference value of node 2 and node 1 is higher than 0.99996 / 0.04791=20.87, which is greater than 1.5, so it is determined that the user 4 and the node 2 exist the correlation relationship;

[0095] S26, the o'2 data set excludes the user 4 related data, and a new to-be-processed data part is selected and formed, and the above process is continued to be processed in circulation;

[0096]

[0097] S27, the next round, confirm that the user 2 data is 5.0360, the Pearson correlation coefficient is-0.0265, 0.1032, 0.3104, 0.7480 respectively, screening determines that user 2 and node 4 exist associated relationship, and continue to cycle processing according to the above process;

[0098] S28, the next round, confirm that the user 1 data is 7.18, the Pearson correlation coefficient is-0.1316, 0.6310, 0.2202, 0.6221 respectively, at this time, the most closely, the second closely monitored point is node 2, 4, because 0.6310 / 0.6221<1.5 relationship is too close, therefore, accumulate the next cycle data, the Pearson correlation coefficient is-0.0170, 0.4375, -0.0950, 0.9640 respectively, 0.9640 / 0.4375>1.5, can determine that user 1 and node 4 exist associated relationship, continue to cycle processing according to the above process;

[0099] S29, the next round, the Pearson correlation coefficient of user 3 is 0.1700, -0.4973, 0.2565, 0.9999 respectively, can determine that user 3 and node 4 exist associated relationship, at this time, the user is confirmed, and the topology structure result is output.

[0100]

[0101] The second embodiment of the application relates to a topology error correction system based on Pearson correlation coefficient and data screening, and specifically comprises a monitoring function module, a correction function module and a storage management module.

[0102] The monitoring function module comprises:

[0103] The first input module is used to obtain the topological relationship matrix of each node in the low-voltage transformer area, and the node comprises a monitoring node and a user node.

[0104] The second input module is used to obtain the power data difference of each node in several rounds, and the power data difference is the difference between the power data collected in the current round and the last round.

[0105] The association module is used to construct the association relationship of the power data difference between nodes based on the topological relationship matrix.

[0106] The screening module is used to calculate the proportion of times when the power data difference of each monitoring node does not satisfy the association relationship, and compare it with the abnormal frequency threshold value, if it exceeds the abnormal frequency threshold value, the current monitoring node is taken as an abnormal monitoring node.

[0107] The correction function module is used to correct the topology structure containing the abnormal monitoring node.

[0108] The topology error correction system is mainly for APP, installed in the intelligent fusion terminal of the transformer district, and runs once every certain period of time to complete the detection function through the use of a plurality of rounds of low-voltage feeder monitoring devices, user electric energy meter power consumption data statistical calculation. The intelligent equipment collection points of the transformer and the branch box are monitoring nodes, and the user intelligent electric energy meter collection points are user nodes. The existing topology is manually initialized and constructed by low-voltage operation and maintenance personnel.

[0109] The topology error correction system specifically realizes the error correction of the topology through the following steps.

[0110] S1, based on the existing topology structure, the monitoring function module is started at a certain time, and the detection result is output.

[0111] Firstly, the correlation algorithm formula based on the power consumption change difference is established as:

[0112] (L o -ΔL o ) T ≤M{[L o L u ]} T ≤(L o +ΔL o ) T

[0113] In the formula,

[0114] L o — the set of power consumption data differences of the monitoring nodes in the current round and the previous round, L o = [l o1 , l o2 , l o3 , …, l oi , …], the higher the current round data is positive, and vice versa;

[0115] ΔL o — the set of power consumption measurement errors of the monitoring nodes, which is a percentage parameter determined by the collection device error;

[0116] L u — the set of power consumption data differences of the user nodes in the current round and the previous round, L u = [l u1 , l u2 , l u3 , …, l uj , …], the higher the current round data is positive, and vice versa;

[0117] ΔL u — the set of power consumption measurement errors of the user nodes, which is a percentage parameter determined by the user electric energy meter error;

[0118] M—topology relationship matrix composed of 1 and 0, a non-0 element indicates that the column number corresponding node and the row number corresponding node have an association relationship, and here the corresponding node represents the monitoring node and the user node arranged in order.

[0119] Then, it is judged whether the data collected by each node meets the association relationship constructed based on the existing topology structure. For L o Set element, representing that the topology relationship related to the node number is abnormal and needs to be output for processing.

[0120] S2, for the abnormal topology relationship, the correction function module is started in time, and the correction is carried out periodically and the corrected topology relationship is output, such as Figure 1 As shown in the figure, the topology correction process includes the following steps:

[0121] Firstly, for the nodes involved in the non-abnormal topology relationship, based on the current data difference set, the abnormal user data part is screened:

[0122]

[0123] In the formula,

[0124] M′—topology relationship matrix M excluding the abnormal part;

[0125] L′ o —abnormal part set of the monitoring node power consumption data difference.

[0126] Secondly, for the abnormal data information, the user data of the specified period is accumulated to form a to-be-calculated data set;

[0127] Thirdly, for the to-be-calculated data set, based on the absolute value, the proportion of the difference value data of each user node in each period is counted, and the user with the largest proportion in the accumulation specified period is screened, that is, the user node with the largest change in the statistical period is screened;

[0128] u=(j|MAX[∑(||l′ uj | / ∑|l′u|)])

[0129] In the formula,

[0130] u—user node number with the largest change in the statistical period;

[0131] l u —user node power consumption data difference value of this round and the last round;

[0132] L′ u —user data set excluding the confirmed correct data.

[0133] Here, ∑|l′ u| is the sum of absolute values of the difference of the user node power consumption in the current period, |l′ uj | / ∑|l′ u | is the difference of the user j power consumption in the current period, ∑(|l′ uj | / ∑|l′ u |) is the sum of the difference of the user j power consumption in the statistical multiple periods, MAX[∑(|l′ uj | / ∑|l′ u |) is the user with the largest difference of the power consumption in the statistical multiple periods.

[0134] Step 4, based on the data set to be calculated, for the data of the user node less than the data of the monitoring node, calculate the Pearson coefficient of the selected user node and the data difference set of the monitoring node, find out the monitoring nodes most closely related to the user node and the second closest monitoring nodes;

[0135] Step 5, compare the correlation of the selected monitoring nodes, for the points with a difference higher than a certain proportion, determine that the user node and the monitoring node have a correlation relationship, and continue to the next step; otherwise, continue to accumulate the next period data outside the current data set, and re-calculate according to the third to fifth steps to determine the correlation relationship;

[0136] Step 6, exclude the determined user node and monitoring node in the data set to be calculated, exclude the correct relationship, and return to step 3 for calculation until all user nodes are calculated.

[0137] S3, according to the corrected topological relationship, perform monitoring in a specified period through the monitoring function module, confirm the correctness of the corrected topological relationship and report.

Claims

1. A low-voltage district topology error correction method, characterized by, The method comprises the following steps: obtaining a topological relationship matrix of each node in a low-voltage transformer area, the nodes including monitoring nodes and user nodes; obtaining power data differences of each node in several rounds, the power data difference being a difference between power data collected in the current round and power data collected in the previous round; based on the topological relationship matrix, constructing an association relationship formula of the power data differences between the nodes; calculating a proportion of times that the power data difference of each monitoring node does not satisfy the association relationship formula, and comparing the proportion of times with an abnormal frequency threshold value, if the proportion of times exceeds the abnormal frequency threshold value, regarding the current monitoring node as an abnormal monitoring node; correcting a topological structure containing the abnormal monitoring node.

2. The method of claim 1, wherein, The step of correcting the topological structure containing the abnormal node comprises: obtaining a user node associated with the abnormal monitoring node as an abnormal user node; screening data associated with the abnormal user node from the collected power data differences to obtain a to-be-calculated data set; extracting an abnormal user node with the largest power data difference variation in a set round in the to-be-calculated data set as a target correction node; according to a correlation coefficient of the target correction node and each abnormal monitoring node, finding an associated monitoring node of the target correction node and updating the to-be-calculated data set according to a matching result, and then returning to the previous step until all abnormal user nodes are matched to the associated monitoring nodes; updating the topological relationship matrix according to each target correction node and the associated monitoring node thereof.

3. The method of claim 2, wherein, The step of finding the associated monitoring node of the target correction node according to the correlation coefficient of the target correction node and each abnormal user node comprises: calculating a Pearson correlation coefficient of the target correction node and each abnormal monitoring node; selecting two abnormal monitoring nodes with the highest and the second highest Pearson correlation coefficients; calculating a ratio of the highest Pearson correlation coefficient to the second highest Pearson correlation coefficient, if the ratio is greater than a set threshold value, regarding the abnormal monitoring node corresponding to the highest Pearson correlation coefficient as the associated monitoring node of the target correction node.

4. The method of claim 2, wherein, The step of updating the to-be-calculated data set according to the matching result comprises: if there is an associated monitoring node, excluding data associated with the target correction node from the to-be-calculated data set; if there is no associated monitoring node, updating the to-be-calculated data set by using the collected power data differences.

5. The method of claim 2, wherein, The to-be-calculated data set is represented as Wherein, L' o is the data set to be calculated, L o is the power data difference of the monitoring node, and M' is the topology relationship matrix M after excluding the abnormal monitoring node and its associated nodes.

6. The method of claim 2, wherein, the target correction node is represented as u = (j | MAX [∑(|l' - j|) / ∑|l' )], l' ∈ L' uj | / ∑|l′ u |)]),l′ u ∈L′ u Wherein, u is a target error correction node, l′ u is the electric quantity data difference of any target error correction node, l′ uj is the electric quantity data difference of the jth target error correction node, L′ u is a to-be-calculated data set.

7. The method of claim 1, wherein, The topological relationship matrix is a matrix composed of 1 and 0, with the monitoring nodes as column nodes and the monitoring nodes and the user nodes as row nodes, wherein a non-0 element indicates that the column node and the row node have an association relationship.

8. The method of claim 1, wherein, The association relationship formula is represented as (L o -ΔL o ) T ≤M{[L o L u ]} T ≤(L o +ΔL o ) T wherein, L o is the power data difference of the monitoring node, ΔL o is the power consumption measurement error of the monitoring node, L u is the power data difference of the user node, ΔL u is the power consumption measurement error of the user node, M is the topological relationship matrix.

9. A low voltage distribution area topology correction system, characterized by, comprises: a first input module, configured to obtain a topological relationship matrix of each node in a low-voltage transformer area, the nodes including monitoring nodes and user nodes; a second input module, configured to obtain power data differences of each node in several rounds, the power data difference being a difference between power data collected in the current round and power data collected in the previous round; an association module, configured to construct an association relationship formula of the power data differences between the nodes based on the topological relationship matrix; The screening module is configured to calculate a proportion of times that the difference between the power data of each monitoring node does not satisfy the correlation formula, and compare the proportion of times with an abnormal frequency threshold value, and if the proportion of times exceeds the abnormal frequency threshold value, the current monitoring node is taken as an abnormal monitoring node; The correction function module is configured to correct the topology structure containing the abnormal monitoring node.

10. The system of claim 9, wherein, The monitoring node includes an intelligent transformer, an intelligent branch box, and an intelligent electrical switch.

Images