A method, system and device for warning of electricity theft behavior by integrating digital and graphical information

Abstract

The present invention belongs to the field of power safety, and specifically relates to a method, system and device for warning of electricity theft behavior by integrating digital and graphical data. The method combines the topological graph of each power metering node in the transformer area with the metering data, and separates the power nodes of commercial users and residential users on the topological graph to obtain a merchant sub-graph and a residential sub-graph. Then, by combining the metering data analysis, the fluctuation similarity and growth similarity of the power consumption data changes of the power users on the two sub-graphs are compared, and the nodes with higher similarity are regarded as safe nodes without the risk of electricity theft. Next, according to the remaining topological graph, each independent risk area is generated, and the corresponding feature sequences are generated by combining the power information of the nodes in the risk area. Finally, using machine learning algorithms, the classification results of the presence of electricity theft behavior are generated for each area according to the input feature sequences. The present invention solves the problems of poor accuracy and insufficient real-time performance existing in the existing electricity theft behavior detection methods based on big data economy.

Description

Technical Field

[0001] The present invention belongs to the field of power safety, and particularly relates to a method, a system and a device for warning of electricity theft behavior by integrating digital and graphical information. Background Art

[0002] Electricity theft refers to the behavior of users stealing electric energy from the power grid through illegal means. On the one hand, electricity theft behavior will cause economic losses to power companies, and on the other hand, it may pose potential safety hazards to the stable operation of the power grid. Although there are various means and types of electricity theft discovered at present, according to the basic working principles adopted by electricity theft behavior, it can be divided into two categories: electricity theft unrelated to metering devices and electricity theft related to metering devices.

[0003] Among them, electricity theft unrelated to metering devices includes users privately increasing capacity, short-circuiting the incoming and outgoing lines of the metering box to steal electricity, etc. Although stealing electricity through these means is relatively concealed and cannot be discovered during daily inspections, the data differences of metering tables at different nodes can be combined to quickly locate the risk areas, and more detailed power inspections can be combined to discover relevant illegal acts. Electricity theft methods related to metering devices include under-voltage electricity theft method, under-current electricity theft method, phase-shifting electricity theft method, differential-expansion electricity theft method, etc. The detection and positioning of this type of electricity theft behavior are relatively difficult, and usually need to be realized through manual detection or analysis methods based on artificial intelligence and big data.

[0004] Currently, the electricity theft detection methods widely used in the process of power grid safety management include two categories: one is the traditional manual detection method, and the other is the method based on big data analysis. The former uses manual sampling to inspect the abnormal electricity consumption caused by electricity theft. However, with the diversification of electricity theft means, the detection difficulty of the manual detection method for electricity theft behavior increases, the probability of finding users with abnormal electricity consumption decreases, and it has a high false alarm rate and missed alarm rate. The latter analyzes the power information collected at each node in the substation area by using machine learning algorithms such as neural networks, and then the big data algorithm identifies and locates the abnormal electricity consumption caused by electricity theft occurring in the substation area according to the operation data of the substation area. The electricity theft behavior detection method based on big data analysis has good generalization ability and can effectively reduce the missed alarm rate. However, the training of machine learning algorithms in such a scheme requires a large amount of abnormal information data of different categories of electricity theft, while in the management center of the power grid, the sample data volume of abnormal information data corresponding to electricity theft behavior is very limited, and the sample data volume corresponding to normal electricity consumption behavior is very large, which causes the model to be prone to falling into local optimum and the convergence speed to slow down during the training stage. In addition, the neural network algorithm needs to process a large amount of power feature information generated by all nodes in the substation area during the actual application stage. In a complex substation area including a large number of power users, this will cause the parameter scale of the algorithm to be very large, increase the requirement of the scheme for computing power, and lead to insufficient real-time performance of the scheme. Summary of the Invention

[0005] To solve the problems existing in the existing detection methods for electricity theft behavior based on big data economy, the present invention provides a method, a system and a device for early warning of electricity theft behavior through digital and graphical fusion.

[0006] The present invention is implemented by adopting the following technical solutions:

[0007] A method for early warning of electricity theft behavior through digital and graphical fusion, which comprises the following steps:

[0008] Obtain the file information and measurement information of all the metering tables in the substation area, and the measurement information includes the daily measurement data and historical measurement data.

[0009] Construct a total topological graph representing the subordination relationship among all the gateway meters, measurement switches and electric energy meters in the substation area in combination with the file information, and divide the total topological graph into a merchant sub-graph and a residential sub-graph according to the types of power users corresponding to each electric energy meter.

[0010] Calculate the per capita electricity consumption increase ICP of each residential user, the daily average electricity consumption growth rate GRD of each commercial user, and the variance of the phase-separated power of each commercial user or residential user respectively according to the daily load data and historical measurement data. 。

[0011] According to , ICP, GRD, generate the feature vectors of each node, calculate the fluctuation similarity S1 and growth similarity S2 between any two electric energy meters in the residential sub-graph or the merchant sub-graph respectively, and merge the nodes with similarity higher than the preset threshold; the electric energy meters of the remaining nodes are recorded as independent meters with the risk of electricity theft.

[0012] Extract the measurement information of each remaining independent meter and its corresponding measurement switch and gateway meter in the merchant sub-graph and the residential sub-graph, generate the feature sequence corresponding to each risk area; and input it into an electricity theft detection model trained based on MLP to generate the classification result of the electricity theft type.

[0013] Based on the classification result output by the electricity theft detection model, conduct a check for electricity theft behavior on the power users corresponding to each independent meter under the measurement switch with the risk of electricity theft.

[0014] As a further improvement of the present invention, the file information includes the device identification number, address information, subordination information of the metering table, and / or user number. The subordination information of the gateway meter includes the device identification number of the measurement switch under its jurisdiction. The subordination information of the measurement switch includes the device identification number of the gateway meter upstream of it and the device identification number of the electric energy meter under its jurisdiction. The subordination information of the electric energy meter includes the device identification number of the measurement switch upstream of it.

[0015] The classification results of the electricity theft types identified by the present invention include: phase-shifting electricity theft, under-voltage electricity theft, differential-extension electricity theft, under-current electricity theft, and no electricity theft.

[0016] As a further improvement of the present invention, the method for splitting the merchant sub-graph is as follows:

[0017] (1) Obtain the total topology graph and the user types of the power users corresponding to each electricity meter included therein. Among them, the total topology graph is a tree graph with each metering table as a node and the subordinate relationship between each metering table as an edge.

[0018] (2) Delete the nodes corresponding to the electricity meters belonging to residential users from the total topology graph.

[0019] (3) Delete the measurement switches that no longer have subordinate electricity meters after the deletion in the above step from the total topology graph.

[0020] (4) Delete the gateway meters that no longer have subordinate range switches after the deletion in the above step from the total topology graph, and the merchant sub-graph can be obtained.

[0021] Correspondingly, by sequentially deleting the nodes corresponding to the electricity meters of the merchant first, and then deleting the nodes corresponding to the range switches and gateway meters without subordinate nodes, the residential sub-graph can be obtained.

[0022] As a further improvement of the present invention, the variance of the split-phase power of commercial users or residential users The calculation formula is:

[0023] ,

[0024] In the above formula, represents the i-phase power of the current node at the t-th sampling on the current day; represents the average value of the i-phase power of the current node on the current day; n represents the daily sampling frequency of the electricity meter.

[0025] As a further improvement of the present invention, the calculation formulas for the increase rate of per capita electricity consumption ICP of residential users and the growth rate of daily average electricity consumption GRD of commercial users are as follows:

[0026] ,

[0027] In the above formula, and respectively represent the cumulative electricity consumption measured by the current electricity meter on the T-th day and the (T-1)-th day; m represents the number of family members of the current residential user.

[0028] As a further improvement of the present invention, the calculation method for the fluctuation similarity S1 between any two nodes a and b is as follows:

[0029] The variance of the three-phase power corresponding to each node 、 、 are used as the spatial coordinates of the node;

[0030] Then, the fluctuation similarity S1 between nodes is the Euclidean distance between the two, and the calculation formula is as follows:

[0031] ,

[0032] In the above formula, x 1 、y 1 and z 1 are the values of the abscissa, ordinate and vertical coordinate of node a respectively; x 2 、y 2 and z 2 are the values of the abscissa, ordinate and vertical coordinate of node b respectively.

[0033] As a further improvement of the present invention, the calculation method of the growth similarity S2 between any two nodes a and b is as follows:

[0034] Encode the ICP or GRP values of nodes a and b within 7 consecutive days into a 7-dimensional feature vector A and B:

[0035] A = (a1, a2, a3, a4, a5, a6, a7),

[0036] B = (a1, a2, a3, a4, a5, a6, a7),

[0037] Then the growth similarity S2 between the two nodes is the cosine distance between the corresponding feature vectors, and the calculation formula is:

[0038] .

[0039] As a further improvement of the present invention, each feature sequence is respectively composed of the feature vectors of the measurement switch and its associated gateway meter and multiple electricity meters, and its data format is as follows:

[0040] ,

[0041] In the above formula, D 1 represents the feature vector of the gateway meter to which the measurement switch belongs; D 2 represents the feature vector of the measurement switch; represents the feature vectors of the 1st to the wth electricity meters under the range switch.

[0042] Among them, the feature vectors of each meter are composed of some or all of the feature values of the following multiple electrical parameters:

[0043] Three-phase voltage U A , U B , U C , three-phase current I A , I B , I C , secondary current I SC , power factor , line loss value L, line loss conversion rate Δ, number of abnormal reverse power Ep.

[0044] The present invention further includes a power theft behavior warning system based on digital and graph fusion, which adopts the power theft behavior warning method based on digital and graph fusion as described above to identify the nodes and risk types with power theft risks in the area. The warning system includes: a topology graph generation module, a sub-graph segmentation module, a similarity calculation unit, a sub-graph merging unit, a feature encoding unit, and a risk prediction unit.

[0045] Among them, the topology graph generation module is used to construct a total topology graph for characterizing the subordination relationship of all metering meters in the area according to the file information of all gateway meters, measurement switches and electric energy meters in the area. The sub-graph segmentation module is used to divide the total topology graph into a merchant sub-graph and a residential sub-graph according to the types of power users corresponding to each electric energy meter.

[0046] The similarity calculation unit is used to calculate the per capita electricity consumption increase ICP of each residential user, the daily average electricity consumption growth rate GRD of each commercial user, and the variance of the split-phase power of each commercial user or residential user respectively according to the load data of the current day and the historical metering data ; and then according to , ICP, GRD to generate the feature vector of each node, and calculate the fluctuation similarity S1 and growth similarity S2 between any two electric energy meters in the residential sub-graph or merchant sub-graph respectively.

[0047] The sub-graph merging unit is used to combine the fluctuation similarity S1 and growth similarity S2 between any two electric energy meters in the residential sub-graph or merchant sub-graph, and merge the nodes corresponding to the electric energy meters with similarity exceeding the preset threshold in the residential sub-graph or merchant sub-graph. The merged nodes are recorded as safe nodes, and the electric energy meters of the remaining nodes are recorded as independent meters with power theft risks.

[0048] The feature encoding unit is used to take the upstream and downstream nodes corresponding to each measurement switch containing an independent meter as risk areas, and generate a feature sequence with the metering information of all metering meters in the risk area as feature values. The risk prediction unit includes a power theft detection model, and the power theft detection model is used to generate and output a classification result for characterizing whether there is a power theft risk in the current area according to the input feature sequence.

[0049] The present invention further includes a digital-graph fusion electricity theft behavior warning device, which includes a memory, a processor, and a computer program stored on the memory and running on the processor. When the processor executes the computer program, the digital-graph fusion electricity theft behavior warning method as described above is implemented, and further, regions with possible electricity theft risks and their corresponding electricity theft types are identified according to the topological relationship and measurement information of all the metering meters in the substation area.

[0050] The technical solution provided by the present invention has the following beneficial effects:

[0051] The present invention provides a digital-graph fusion electricity theft behavior warning method, which combines the topological graph of each power metering node in the substation area with the measurement data, and separates the power nodes of commercial users and residential users on the topological graph to obtain two subgraphs. Then, the similarity of the changes in the electricity consumption data of the power users on the two subgraphs is analyzed in combination with the measurement data, and the nodes with higher similarity are regarded as safe nodes without electricity theft risks. Next, characteristic sequences corresponding to the power data of the nodes in each independent risk region in the remaining topological graph are generated. Finally, a machine learning algorithm is used to generate classification results indicating the existence of electricity theft behavior in each region according to the input characteristic sequences.

[0052] The solution provided by the present invention separately identifies the electricity theft risks of different types of power users, and pre-analyzes the security and merges the nodes in the sub-topological graphs of various power users in combination with the actual situation. This can effectively remove the interference information contained in the characteristic information of each node in the original substation area, and further greatly reduce the dimension of the input features of the machine learning algorithm, making the training difficulty of the network model greatly reduced, improving the dependence of the classification accuracy of the model on the sample data scale, and enhancing the classification accuracy and generalization of the network model.

[0053] In addition, the parameter scale of the network model adopted by the solution of the present invention is smaller, which on the one hand improves the real-time performance of the solution, and on the other hand enables it to be deployed in existing embedded devices (such as intelligent fusion terminals) in the substation area, improving the practical value of the solution. Description of the Drawings

[0054] Figure 1 It is a flowchart of the steps of the digital-graph fusion electricity theft behavior warning method provided in Embodiment 1 of the present invention.

[0055] Figure 2 It is a typical total topological graph of the substation area constructed in Embodiment 1 of the present invention.

[0056] Figure 3 It is a flowchart of the steps of the segmentation method of the merchant subgraph in Embodiment 1 of the present invention.

[0057] Figure 4It is the flowchart of the steps of the segmentation method for the residential sub - graph in Embodiment 1 of the present invention.

[0058] Figure 5 It is Figure 2 the merchant sub - graph segmented from the total topological graph of

[0059] Figure 6 It is Figure 2 the residential sub - graph segmented from the total topological graph of

[0060] Figure 7 It is the flowchart of the steps of the calculation method for the fluctuation similarity between any two nodes in Embodiment 1 of the present invention.

[0061] Figure 8 It is the flowchart of the steps of the calculation method for the growth similarity between any two nodes in Embodiment 1 of the present invention.

[0062] Figure 9 It is Figure 5 the distribution map of the risk areas included in the merchant sub - graph shown in

[0063] Figure 10 It is the system architecture diagram of the power - stealing behavior warning system for digital - graph fusion provided in Embodiment 2 of the present invention. Detailed implementation manners

[0064] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0065] Embodiment 1

[0066] This embodiment provides a power - stealing behavior warning method for digital - graph fusion. The technical idea of this solution is as follows: First, construct a topological graph containing all the nodes corresponding to the metering meters in the transformer area, and use the metering information of each metering meter as the feature information of the node. Then, according to the type of each power user, divide the overall topological graph of the transformer area into two sub - graphs, one only containing residential users and the other only containing commercial users. Next, analyze the fluctuation and growth conditions of the historical power consumption data of each power user in the two sub - graphs, and divide the nodes with similar fluctuation and growth conditions into safe nodes, and the remaining nodes are divided into risk nodes that may have power - stealing behaviors. Finally, extract the metering data of the electric energy meters of these risk nodes and their upstream measuring switches and gateway meters, encode them into feature sequences, and classify the input feature sequences through a power - stealing analysis model based on machine - learning algorithms to determine whether there is a power - stealing behavior and the corresponding power - stealing type of the power - stealing behavior.

[0067] Combined with the technical concept of the embodiment described above, it can be seen that, unlike the traditional solution that directly uses the power information obtained by all nodes to detect power theft, the embodiment discusses the two separately according to the different rules of residential power consumption and commercial power consumption, and before inputting the node information into the machine learning algorithm, it predicts whether the node has the risk of power theft based on the changes in the power consumption data of each power user, and then only inputs the power information related to the node with the risk of power theft into the machine learning algorithm. This new data processing strategy can greatly reduce the computational load of the machine learning algorithm and reduce the algorithm's dependence on sample data during the training phase. On the basis of ensuring the final prediction accuracy, it improves the real-time performance of the algorithm and enables the power theft warning solution to be deployed in embedded devices in the substation area.

[0068] Specifically, Figure 1 As shown, the electricity theft warning method based on data-graph fusion provided in this embodiment includes the following steps:

[0069] S1: Obtain the archive information and metering information of all meters in the substation. The metering information includes the metering data of the day and the historical metering data.

[0070] In this embodiment, the meters installed in the substation include three types, namely, gateway meters, measuring switches and electric energy meters. Among them, the electric energy meter is installed on the user side to measure the electricity consumption information of each power user. The measuring switch is installed at a branch node of the power supply line and serves as the master meter of several electric energy meters under its jurisdiction. The gateway meter is installed on the main line of the substation and serves as the master meter of several range switches under its jurisdiction.

[0071] For the above three types of meters, the archive information includes the device identification number, address information, subordinate information, and / or user number of the meter. Among them, the device identification number is the exclusive identification code of each meter; the address information corresponds to the installation location of the meter at the node; for the electric energy meter, its archive information also includes the user number, which is a unique account number obtained by each power user when handling the power supply business at the power company. By querying the business database of the power company in combination with the account number, the user type, historical electricity consumption data and member information of each user can also be obtained. Subordinate information refers to the subordinate relationship between each meter in the substation and its upstream or subordinate meters. Specifically, in the topological structure of each meter in the substation, the gateway table is located at the top level, the measuring switch is located at the middle level, and the electric energy meter is located at the bottom level. Therefore, the subordinate information of the gateway table includes the device identification number of the measuring switch under its jurisdiction; the subordinate information of the measuring switch includes the device identification number of the upstream gateway table and the device identification number of the electric energy meter under its jurisdiction; the subordinate information of the electric energy meter includes the device identification number of the upstream measuring switch.

[0072] In this embodiment, the metering information includes a series of power characteristic parameters that can be collected or indirectly calculated between meters, such as: three-phase voltage U A 、U B 、U C , three-phase current I A 、I B 、I C , secondary current I SC , power factor , line loss value L, line loss conversion rate Δ, number of abnormal reverse power Ep, and so on.

[0073] S2: According to the three-level pyramid structure of the gateway meter, measurement switch, and electricity meter, combined with the archive information, construct a total topology diagram representing the subordination relationship of all meters in the substation area.

[0074] In this embodiment, the total topology diagram of the substation area can be generated through the subordination information of each meter in the substation area. The total topology diagram is a tree diagram with each meter as a node and the subordination relationship between each meter as an edge. A typical total topology diagram includes three layers. The number of gateway meters in the top layer is the least, the number of measurement switches is the second, and the data volume of electricity meters in the bottom layer is the largest. As Figure 2 shown, the nodes corresponding to the three types of meters in the total topology form a typical pyramid structure.

[0075] S3: According to the types of power users corresponding to each electricity meter, divide the total topology diagram into a merchant sub-diagram and a residential sub-diagram.

[0076] In the solution of this embodiment, the technical personnel found that the statistical characteristics of the electricity consumption data of different types of power users are significantly different. Therefore, when analyzing the electricity theft behavior of different nodes, different characteristic information should be extracted in combination with different types of power users. Therefore, in this embodiment, different types of user types in the substation area are divided into different sub-diagrams, and the electricity theft behavior of each power user included in each sub-diagram is analyzed, which can effectively improve the classification accuracy of the solution and reduce the parameter scale of the trained machine learning algorithm.

[0077] Specifically, as Figure 3 shown, the method for dividing the merchant sub-diagram is as follows:

[0078] (1) Obtain the total topology diagram and the user types of the power users corresponding to each electricity meter included therein.

[0079] (2) Delete the nodes corresponding to the electricity meters belonging to residential users from the total topology diagram.

[0080] (3) Delete the measurement switches that no longer have subordinate electricity meters after the deletion in the previous step from the total topology diagram.

[0081] (4) Delete the gateway meter without a subordinate range switch after the above step from the total topology diagram, and the merchant sub-diagram can be obtained.

[0082] Correspondingly, as Figure 4 shown, the splitting method of the residential sub-diagram is as follows:

[0083] (1) Obtain the total topology diagram and the user types of the power users corresponding to each electricity meter included therein.

[0084] (2) Delete the nodes corresponding to the electricity meters belonging to commercial users from the total topology diagram.

[0085] (3) Delete the measurement switches that no longer have subordinate electricity meters after the deletion in the above step from the total topology diagram.

[0086] (4) Delete the gateway meter without a subordinate range switch after the deletion in the above step from the total topology diagram, and the merchant sub-diagram can be obtained.

[0087] For example, based on the splitting methods of the merchant sub-diagram and the residential sub-diagram introduced above, split the Figure 2 shown typical total topology diagram, and the merchant sub-diagram as shown in Figure 5 and the residential sub-diagram as shown in Figure 6 can be obtained.

[0088] S4: Calculate the per capita electricity consumption increase ICP of each residential user, the daily average electricity consumption growth rate GRD of each commercial user, and the variance of the split-phase power of each commercial user or residential user respectively according to the load data of the current day and the historical metering data .

[0089] Considering that in the real transformer area scenario, each commercial sub-diagram or residential sub-diagram still contains a large number of power user nodes, and most of these power user nodes actually do not have electricity theft behavior. Therefore, in this embodiment, the safe nodes that obviously do not have electricity theft behavior are excluded by combining the residential sub-diagram and the commercial sub-diagram.

[0090] In the process of designing the criterion for excluding safe nodes, the technical personnel in this case found that for ordinary residential users, the change in their daily electricity consumption will be affected by natural climate and daily activities, and shows obvious statistical laws in the time domain. For example, in the high-temperature period in summer and the low-temperature period in winter, due to the increased demand for refrigeration and heating, the electricity consumption of ordinary residential users will increase significantly compared with other periods. In addition, the electricity consumption data of residential users is relatively stable on rest days, while on weekdays, the electricity consumption of residential users will increase significantly after family members get off work or school, etc.

[0091] Theoretically speaking, combining the time-domain variation of the electricity consumption of users over a relatively long period will help analyze whether there are abnormalities in the electricity consumption data of users. However, this data analysis strategy involves a large amount of data processing and is not suitable for practical applications. In view of this situation, the technical personnel in this embodiment have found a new way. They first consider analyzing the fluctuations and growth of the electricity consumption data of users of the same type, then analyze the similarity of the fluctuations and growth of the electricity consumption of power users at different nodes during the same period, and finally combine the similarity of the changes in the electricity consumption data between nodes to determine whether there is a risk of electricity theft at the corresponding nodes.

[0092] Taking residential users as an example, if the fluctuations and growth of the electricity consumption of any two users are similar during a certain period, it indicates that the changes in their electricity consumption data are reasonable, and the changes in electricity consumption may be caused by the local natural climate or the laws of group activities. At this time, both of these two nodes can be considered safe nodes. On the contrary, when the differences in the fluctuations and growth of the electricity consumption of any two users are large during a certain period, it indicates that the changes in their electricity consumption data may not be reasonable, and this data difference may be due to the inconsistency between the measured electricity consumption and the actual electricity consumption of one of them, that is, there is an electricity theft behavior. At this time, these two nodes should be considered as nodes with a risk of electricity theft.

[0093] Correspondingly, for commercial users, the prosperity levels of the commercial activities of various merchants in the same region are usually similar; and there is often a positive correlation between the prosperity level of commercial activities and the electricity consumption of merchants. Therefore, in this embodiment, first analyze the fluctuations and growth of the electricity consumption data of each commercial user, and then analyze whether the fluctuations and growth of the electricity consumption of power users at different nodes are similar during the same period. If they are, they are determined as safe nodes; otherwise, they are regarded as risk nodes.

[0094] Specifically, in this embodiment, the variance of the split-phase power and the increase rate of per capita electricity consumption ICP are respectively set as indicators to evaluate the fluctuations and growth of the daily electricity consumption of residential users, and the variance of the split-phase power and the daily average electricity consumption growth rate GRD are set as indicators to evaluate the fluctuations and growth of the daily electricity consumption of commercial users.

[0095] Among them, the variance of the split-phase power of commercial users and residential users is calculated using the same formula, and the calculation formula is as follows:

[0096] ,

[0097] In the above formula, represents the i-phase power of the current node at the t-th sampling on the current day; It represents the average power value of the i-phase on the current day of the current node; n represents the daily sampling frequency of the meter.

[0098] Among them, the per capita electricity consumption increase ICP of residential users refers to the difference between the per capita electricity consumption of residential users on the current day and that on the previous day. The daily average electricity consumption growth rate GRD of commercial users refers to the average value of the year-on-year daily growth rate of electricity consumption of commercial users within 30 consecutive days. The calculation formulas of ICP and GRD are as follows:

[0099] ,

[0100] In the above formula, and respectively represent the cumulative electricity consumption measured by the current meter on the Tth day and the (T - 1)th day; m represents the number of family members of the current residential user.

[0101] S5: Generate the feature vector of each node according to , ICP, and GRD, calculate the fluctuation similarity S1 and growth similarity S2 between any two electricity meters in the residential subgraph or merchant subgraph respectively, and merge the nodes with similarity higher than the preset threshold; the electricity meters of the remaining nodes are recorded as independent meters with electricity theft risks.

[0102] In this embodiment, as Figure 7 shown, the calculation method of the fluctuation similarity S1 between any two nodes a and b is as follows:

[0103] Take the variances of the three-phase power corresponding to each node , , as the spatial coordinates of this node, then there is:

[0104] The fluctuation similarity S1 between nodes is the Euclidean distance between them, and the calculation formula is as follows:

[0105] ,

[0106] In the above formula, x 1 , y 1 and z 1 are the values of the abscissa, ordinate, and vertical coordinate of node a respectively; x 2 , y 2 and z 2 are the values of the abscissa, ordinate, and vertical coordinate of node b respectively.

[0107] As Figure 8 shown, the calculation method of the growth similarity S2 between any two nodes a and b is as follows:

[0108] Encode the values of ICP or GRP of nodes a and b within 7 consecutive days into a 7-dimensional feature vector A and B:

[0109] A = (a1, a2, a3, a4, a5, a6, a7),

[0110] B = (a1, a2, a3, a4, a5, a6, a7),

[0111] Then the growth similarity S2 between the two nodes is the cosine distance between the corresponding feature vectors, and the calculation formula is:

[0112] .

[0113] In this embodiment, first calculate the fluctuation similarity S1 and the growth similarity S2 between any two electric energy meters in the residential sub - graph or the merchant sub - graph. Next, first perform the first - round merging of the nodes corresponding to each electric energy meter according to S1, and then perform the second - round merging of the nodes corresponding to each electric energy meter according to S2. The operation strategies for the two - round merging are as follows:

[0114] First, classify each node with S1 or S2 greater than or equal to the preset similarity threshold into the safe node set, and then classify each node with S1 or S2 less than the preset similarity threshold into the risk node set. Finally, considering that among the similarity results of any two nodes calculated, some nodes have a relatively high similarity with some nodes (safe nodes) and a relatively low similarity with other nodes (risk nodes), these nodes should be regarded as safe nodes according to the actual situation. That is: remove the nodes that appear in both the safe node set and the risk node set from the risk node set.

[0115] S6: Extract the measurement information of each remaining independent electric meter and its corresponding measurement switch and gateway meter in the merchant sub - graph and the residential sub - graph, and generate a feature sequence corresponding to each risk area. Input each feature sequence into a power - stealing detection model trained based on MLP to generate a classification result of the power - stealing type.

[0116] After the two - round merging of safe nodes in step S5, only a small number of risk nodes remain in the residential sub - graph or the merchant sub - graph. On this basis, as Figure 9 shown, this step defines the sub - graph area composed of the same measurement switch and its associated gateway meter and multiple independent electric meters in the residential sub - graph or the merchant sub - graph as a risk area, and each residential sub - graph or merchant sub - graph may contain multiple risk areas. Next, this embodiment extracts the measurement information of all the meters in each risk area and encodes the measurement information into a feature sequence.

[0117] Specifically, each feature sequence is respectively composed of the feature vectors of the measurement switch and its associated gateway meter and multiple electric energy meters, and its data format is as follows:

[0118] ,

[0119] In the above formula, D 1 represents the eigenvector of the gateway meter to which the measurement switch belongs; D 2 represents the eigenvector of the measurement switch; represents the eigenvectors of the first to the wth watt-hour meters under the range switch.

[0120] Among them, the eigenvector of each meter is composed of a series of eigenvalue of power parameters. The parameters included in the eigenvector of each type of meter are selected from the following power characteristic parameters:

[0121] Three-phase voltages U A , U B , U C , three-phase currents I A , I B , I C , secondary current I SC , power factor , line loss value L, line loss conversion rate Δ, number of abnormal reverse power Ep.

[0122] It should be additionally noted that for any type of meter, its eigenvector can select all of the above power characteristic parameters at the same time, or can select a part of the above power characteristic parameters.

[0123] The characteristic sequences corresponding to the generated risk regions will be input into a pre-trained electricity theft detection model, which is used to generate the classification results of the corresponding electricity theft types according to the input characteristic sequences. In this embodiment, the classification results of the electricity theft types identified by the model include five types, namely: electricity theft by phase-shifting method, electricity theft by under-voltage method, electricity theft by differential expansion method, electricity theft by under-current method, and no electricity theft. Among them, "no electricity theft" means that there is no electricity theft risk predicted by the model in this risk region.

[0124] In a more optimized solution of this embodiment, in order to eliminate the interference of the dimensions of different power characteristic parameters on the prediction performance of the electricity theft detection model, it is necessary to normalize the value of each characteristic parameter included in the generated characteristic sequence. The calculation formula for normalization is as follows:

[0125] ,

[0126] In the above formula, x represents the original value of the characteristic parameter; represents the normalized result of the characteristic parameter; and represent the maximum value and the minimum value of the characteristic parameter respectively.

[0127] In addition, it should be noted that: the solution provided in this embodiment can use the same trained electricity theft detection model to classify the feature sequences in the two subgraphs at the same time, or can use a large number of feature sequences from the risk areas in the commercial subgraph and the residential subgraph to train a corresponding electricity theft detection module respectively, and then use the two electricity theft detection modules to analyze the electricity theft risks of the feature sequences from different sources respectively.

[0128] S8: Generate a warning message based on the classification result output by the electricity theft detection model, and conduct a screening for electricity theft behaviors of the power users corresponding to each independent electricity meter under the measurement switch with the risk of electricity theft.

[0129] In the solution of this embodiment, the classification result output by the electricity theft detection model in the previous step can reflect whether there is an electricity theft behavior in the risk area and the type of the corresponding electricity theft behavior. In this step, the management center of the power grid can generate a warning message according to the corresponding classification result, and dispatch a work order to the operation and maintenance personnel in this area. Relevant personnel arrive at the scene to conduct a manual inspection of the metering tables of each power user, find the object implementing the electricity theft behavior, solidify the evidence and initiate subsequent legal procedures to deal with the illegal behavior.

[0130] Embodiment 2

[0131] Based on the solution of Embodiment 1, as Figure 2 shown, this embodiment further provides a digital graph fusion-based electricity theft behavior warning system, which is a computer program for implementing the solution of Embodiment 1, and can be stored in a storage medium or a computer device. Furthermore, when the computer program runs, it realizes the digital graph fusion-based electricity theft behavior warning method in Embodiment 1, and identifies the nodes and risk types with the risk of electricity theft in the area according to the monitoring data generated in real time in the area. As Figure 10 shown, the warning system includes: a topology graph generation module, a subgraph segmentation module, a similarity calculation unit, a subgraph merging unit, a feature encoding unit, and a risk prediction unit.

[0132] Among them, the topology graph generation module is used to construct a total topology graph for representing the subordinate relationship of all metering tables in the area according to the file information of all gateway meters, measurement switches and electricity meters in the area. The subgraph segmentation module is used to divide the total topology graph into a commercial subgraph and a residential subgraph according to the type of the power user corresponding to each electricity meter.

[0133] The similarity calculation unit is used to calculate the per capita electricity consumption increase ICP of each residential user, the daily average electricity consumption growth rate GRD of each commercial user, and the variance of the split-phase power of each commercial user or residential user respectively according to the load data of the current day and the historical metering data ; and then according to , ICP and GRD generate the eigenvectors of each node, and calculate the fluctuation similarity S1 and growth similarity S2 of any two electricity meters in the residential subgraph or merchant subgraph respectively.

[0134] The subgraph merging unit is used to combine the fluctuation similarity S1 and growth similarity S2 of any two electricity meters in the residential subgraph or merchant subgraph, and merge the nodes corresponding to the electricity meters with any similarity exceeding the preset threshold in the residential subgraph or merchant subgraph. The merged nodes are recorded as safe nodes, and the electricity meters of the remaining nodes are recorded as independent meters with electricity theft risks.

[0135] The feature encoding unit is used to take the upstream and downstream nodes corresponding to each measurement switch containing an independent meter as risk areas, and generate a feature sequence with the measurement information of all the meters in the risk area as eigenvalues. The risk prediction unit includes an electricity theft detection model, which is used to generate and output a classification result for characterizing whether there is an electricity theft risk in the current area according to the input feature sequence.

[0136] Embodiment 3

[0137] This embodiment provides an electricity theft behavior warning device based on digital graph fusion, which includes a memory, a processor, and a computer program stored on the memory and running on the processor. When the processor executes the computer program, it implements the electricity theft behavior warning method based on digital graph fusion in Embodiment 1, and further realizes identifying the areas that may have electricity theft risks and their corresponding electricity theft types according to the topological relationship and measurement information of all the meters in the substation area.

[0138] The electricity theft behavior warning device based on digital graph fusion provided in this embodiment is essentially a computer device. In the actual application process, this computer device can adopt an embedded device and be deployed in a concentrator or fusion terminal with strong data processing capabilities in the station. It can also sample an independent computer device, such as a laptop, a tablet computer, a desktop computer, or a medium and large computer device such as a rack-mounted server, a blade server, a tower server, or a cabinet server (including an independent server, or a server cluster composed of multiple servers) that can execute computer programs.

[0139] The computer device of this embodiment at least includes, but is not limited to, a memory and a processor that can communicate with each other through a system bus. In this embodiment, the memory (i.e., the readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory can be an internal storage unit of the computer device, such as the hard disk or memory of the computer device. In other embodiments, the memory can also be an external storage device of the computer device, such as a plug-in hard disk equipped on the computer device, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Of course, the memory can also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the memory is usually used to store the operating system installed on the computer device and various application software. In addition, the memory can also be used to temporarily store various data that have been output or will be output.

[0140] In some embodiments, the processor can be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor is usually used to control the overall operation of the computer device.

[0141] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A digital-graphic fusion electricity theft behavior early warning method, characterized in that: It includes the following steps: Obtain archive information and metering information of all meters in the substation area, including daily metering data and historical metering data; Combined with archival information, a general topological diagram is constructed to represent the subordinate relationships among all gateway meters, measuring switches and electric energy meters in the substation area; and dividing the total topology map into a business sub-map and a resident sub-map according to the type of power user corresponding to each power meter; Based on the load data of the day and the historical metering data, the ICP of the per capita electricity consumption of each residential user, the GRD of the daily average electricity consumption of each commercial user, and the variance of the phase power of each commercial user or residential user are calculated respectively. ; according to , ICP, GRD generate the feature vector of each node, calculate the fluctuation similarity S1 and growth similarity S2 of any two electric energy meters in the resident subgraph or merchant subgraph, and merge the nodes with similarity higher than the preset threshold; the electric energy meters of the remaining nodes are recorded as independent electric energy meters with the risk of electricity theft; Extracting metering information of each remaining independent electric meter and its corresponding measuring switch and gateway meter in the merchant sub-map and the resident sub-map, and generating a feature sequence corresponding to each risk area; And input it into an electricity theft detection model trained based on MLP to generate a classification result of electricity theft type; Based on the classification result output by the electricity theft detection model, the electricity theft behavior of the corresponding power users of each independent electricity meter under the measuring switch with the risk of electricity theft is checked.

2. The method for early warning of electricity theft based on digital-graphic fusion as claimed in claim 1, characterized in that: The file information includes the device identification number, address information, subordinate information, and / or user number of the meter; the subordinate information of the gateway meter includes the device identification number of the measuring switch under it; the subordinate information of the measuring switch includes the device identification number of the upstream gateway meter and the device identification number of the electric energy meter under it; the subordinate information of the electric energy meter includes the device identification number of the upstream measuring switch; The classification results of the electricity theft types include: phase shift electricity theft, undervoltage electricity theft, differential expansion electricity theft, undercurrent electricity theft and no electricity theft.

3. The method for early warning of electricity theft based on digital-graphic fusion as claimed in claim 2, characterized in that: The method for segmenting the merchant subgraph is as follows: (1) Obtaining the overall topology map and the user type of the power user corresponding to each electric energy meter contained therein; The overall topology diagram is a tree diagram with each meter as a node and the subordinate relationship between each meter as an edge; (2) deleting the nodes corresponding to the electric energy meters belonging to the residential users from the overall topology map; (3) Deleting the measuring switches that no longer have subordinate electric energy meters after the above step is deleted from the overall topology diagram; (4) Deleting the gateway table that no longer has the range switches under it after the deletion in the previous step from the overall topology map, thereby obtaining the merchant sub-map; Accordingly, the nodes corresponding to the electricity meters of the merchants are deleted first, and then the nodes corresponding to the range switches and gateway meters without subordinate nodes are deleted, so as to obtain the resident subgraph.

4. The method for early warning of electricity theft based on digital-graphic fusion as claimed in claim 3, characterized in that: The variance of the phase power of commercial or residential users The calculation formula is: ， In the above formula, Indicates the i-phase power of the current node at the t-th sampling time of the day; It represents the average power of phase i of the current node on that day; n represents the daily sampling frequency of the meter.

5. The method for early warning of electricity theft based on digital-graphic fusion as claimed in claim 4, characterized in that: The calculation formula for the ICP of the per capita electricity consumption of residential users and the GRD of the average daily electricity consumption of commercial users is as follows: , In the above formula, and They represent the cumulative electricity volume measured by the current meter on day T and day T-1 respectively; m represents the number of people in the current household of the residential user.

6. The method for early warning of electricity theft based on digital-graphic fusion as claimed in claim 5, characterized in that: The calculation method of the fluctuation similarity S1 between any two nodes a and b is as follows: The variance of the three-phase power corresponding to each node , , As the spatial coordinates of the node; Then, the fluctuation similarity S1 between nodes is the Euclidean distance between them, and the calculation formula is as follows: ， In the above formula, x 1 ,y 1 and z 1 are the values ​​of the horizontal, vertical and vertical coordinates of node a respectively; x 2 ,y 2 and z 2 are the values ​​of the horizontal, vertical and vertical coordinates of node b respectively.

7. The method for early warning of electricity theft based on digital-graphic fusion as claimed in claim 1, characterized in that: The calculation method of the growth similarity S2 of any two nodes a and b is as follows: Encode the ICP or GRP values ​​of nodes a and b in 7 consecutive days into a 7-dimensional feature vector A and B: A=(a1, a2, a3, a4, a5, a6, a7), B=(a1, a2, a3, a4, a5, a6, a7), Then the growth similarity S2 of two nodes is the cosine distance between the corresponding feature vectors, and the calculation formula is: 。 8. The method for early warning of electricity theft based on digital-graphic fusion as claimed in claim 7, characterized in that: Each feature sequence is composed of the feature vectors of the measurement switch and its associated gateway table and multiple electric energy meters, and its data format is as follows: ， The above formula, D 1 The characteristic vector representing the gateway table to which the measurement switch belongs; i>D 2 The characteristic vector representing the measurement switch; Then it represents the characteristic vector of the 1st to wth electric energy meters under the range switch; The characteristic vector of each meter is composed of part or all of the characteristic values ​​of the following multiple power parameters: Three-phase voltage U A , U B , U C , three-phase current I A ,I B ,I C , secondary current I SC , power factor , line loss value L, line loss conversion rate Δ, number of reverse power abnormalities Ep.

9. A digital-graphic fusion electricity theft warning system, characterized in that: It adopts the electricity theft behavior early warning method of digital-graph fusion as described in any one of claims 1 to 8 to identify nodes and risk types with electricity theft risks in the test area; The electricity theft warning system comprises: A topology map generation module is used to construct a general topology map for representing the subordinate relationship of all meters in the substation area based on the archive information of all the gateway meters, measuring switches and electric energy meters in the substation area; A subgraph segmentation module, which is used to segment the general topology graph into a merchant subgraph and a resident subgraph according to the type of power user corresponding to each electric energy meter; The similarity calculation unit is used to calculate the per capita electricity consumption increase ICP of each residential user, the daily average electricity consumption growth rate GRD of each commercial user, and the variance of the phase power of each commercial user or residential user based on the load data of the day and the historical metering data. ; Then according to , ICP, GRD generate the feature vector of each node, and calculate the fluctuation similarity S1 and growth similarity S2 of any two electric energy meters in the resident subgraph or merchant subgraph respectively; A subgraph merging unit is used to combine the fluctuation similarity S1 and growth similarity S2 of any two electric energy meters in the resident subgraph or the merchant subgraph, merge the nodes corresponding to any electric energy meters in the resident subgraph or the merchant subgraph whose similarity exceeds a preset threshold, and record the merged nodes as safe nodes, and record the electric energy meters of the remaining nodes as independent electric energy meters with the risk of electricity theft; A feature encoding unit, which is used to take the upstream and downstream nodes corresponding to each measuring switch containing an independent electric meter as a risk area, and generate a feature sequence with the metering information of all meters in the risk area as feature values; The risk prediction unit includes an electricity theft detection model, which is used to generate and output a classification result for characterizing whether there is an electricity theft risk in the current area based on an input feature sequence.

10. A digital-graphic fusion electricity theft warning device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, it implements the method for early warning of electricity theft by integrating data and graph as described in any one of claims 1 to 8, thereby identifying areas where there may be risks of electricity theft and their corresponding types of electricity theft based on the topological relationship and metering information of all meters in the substation.

Images