A method and device for monitoring the wiring of an electric energy meter, an electronic device and a storage medium

Abstract

The application provides a kind of electric energy meter wiring monitoring method, device, electronic equipment and storage medium, it is related to electric energy meter technical field, method includes obtaining the basic attribute data and historical power consumption data of target electric energy meter;Based on the complete state of the historical power consumption data of target electric energy meter, corresponding comparison electric energy meter set is determined;Based on the basic attribute data of target electric energy meter, first comparison electric energy meter is matched in comparison electric energy meter set;Based on the historical power consumption data of target electric energy meter and the historical power consumption data of each first comparison electric energy meter, multi-dimension score is carried out, to obtain the dynamic credibility value of each first comparison electric energy meter;Based on the dynamic credibility value of each first comparison electric energy meter, sorting is carried out, to determine second comparison electric energy meter;The historical power consumption data corresponding to target electric energy meter and second comparison electric energy meter is input into wiring state prediction model, to obtain wiring state prediction result, wiring state prediction result is used to indicate whether the wiring of target electric energy meter is abnormal.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric energy meter, in particular to a monitoring method and device for electric energy meter wiring, an electronic device and a storage medium. BACKGROUND

[0002] With the rapid development of smart grid and Internet of Things technology, remote data acquisition has been widely used in electric energy metering field. The wiring error detection scheme based on remote data acquisition usually adopts a reference comparison method, that is, by comparing and analyzing the data of the target electric energy meter and the reference electric energy meter, it is determined whether the target meter has wiring error. This scheme requires that the reference electric energy meter should be consistent with the target meter in terms of key parameters such as "wiring type, protocol type, user type", so as to ensure the reliability of the comparison data.

[0003] However, in the actual operation environment of the power grid, even the same type of electric energy meter may have different data acquisition results due to different equipment aging degrees, load characteristics differences, installation environment changes and other factors, which brings challenges to the wiring error detection work. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a monitoring method and device for electric energy meter wiring, an electronic device and a storage medium, to solve the technical problem that the reference electric energy meter in the electric energy meter remote wiring detection scheme is not accurate, which affects the accuracy of wiring abnormality detection.

[0005] In a first aspect, the present application provides a monitoring method for electric energy meter wiring, which comprises performing the following steps on a target electric energy meter in a current area according to a preset time interval to determine whether the target electric energy meter has wiring abnormality:

[0006] obtaining the basic attribute data and the historical power consumption data of the target electric energy meter; based on the completeness of the historical power consumption data of the target electric energy meter, determining a corresponding comparison electric energy meter set, which is a standard comparison electric energy meter set or a temporary comparison electric energy meter set; matching a first comparison electric energy meter in the comparison electric energy meter set based on the basic attribute data of the target electric energy meter; performing multi-dimension scoring based on the historical power consumption data of the target electric energy meter and the historical power consumption data of each first comparison electric energy meter to obtain a dynamic credibility value of each first comparison electric energy meter; sorting based on the dynamic credibility value of each first comparison electric energy meter to determine a second comparison electric energy meter; inputting the historical power consumption data corresponding to the target electric energy meter and the second comparison electric energy meter into a wiring state prediction model to obtain a wiring state prediction result, which is used to indicate whether the wiring of the target electric energy meter is abnormal.

[0007] In an optional embodiment, the basic attribute data at least includes an attribute identifier, the attribute identifier includes a plurality of basic attribute fields, each basic attribute field indicates one of wiring mode, communication protocol type, user type, device model, and production batch, and the first comparison electric energy meter is matched in the following manner:

[0008] For each electric energy meter in the comparison electric energy meter set, the attribute identifier of the electric energy meter is matched with the attribute identifier of the target electric energy meter in the basic attribute field;

[0009] If all the basic attribute fields are the same, it is determined that the electric energy meter is the first comparison electric energy meter.

[0010] In an optional embodiment, the dynamic confidence value of each first comparison electric energy meter is calculated in the following manner:

[0011] Based on the historical electricity consumption data of the target electric energy meter / first comparison electric energy meter, daily load curves, monthly load curves, daily load rate interval curves, and electricity fluctuation amount curves are respectively determined;

[0012] The distance values corresponding to the daily load curves, monthly load curves, daily load rate interval curves, and electricity fluctuation amount curves of the target electric energy meter and the first comparison electric energy meter are respectively calculated by the dynamic time warping algorithm;

[0013] Based on the calculated multiple distance values and corresponding weight factors, the dynamic confidence value of the first comparison electric energy meter is weightedly calculated.

[0014] In an optional embodiment, the target electric energy corresponding to the standard comparison electric energy meter set and the temporary comparison electric energy meter set are divided in the following manner:

[0015] For all electric energy meters in the current area that have the same feeder and transformer as the target electric energy meter, it is determined whether the use duration of the electric energy meter is greater than a preset duration;

[0016] If yes, it is determined whether the electric energy meter has a fault record, and if no, the electric energy meter is added to the temporary comparison electric energy meter set;

[0017] If not, it is determined whether the electricity consumption data uploaded by the electric energy meter is stable, and if so, the electric energy meter is added to the temporary comparison electric energy meter set;

[0018] If stable, the electric energy meter is added to the standard comparison electric energy meter set, and if unstable, the electric energy meter is added to the temporary comparison electric energy meter set.

[0019] In an optional embodiment, based on the completeness of the historical electricity consumption data of the target electric energy meter, the steps of determining the corresponding comparison electric energy meter set include:

[0020] determine whether the use duration of the target electric energy meter is greater than a preset duration, and determine whether the power consumption data uploaded by the target electric energy meter is stable;

[0021] If both conditions are met, the standard comparison electric energy meter set is taken as the comparison electric energy meter set, otherwise, the temporary comparison electric energy meter set is taken as the comparison electric energy meter set.

[0022] In an optional embodiment, before the step of sorting based on the dynamic credibility values of the first comparison electric energy meters, the method further comprises:

[0023] Based on the on-site investigation results of the electric energy meters, the dynamic credibility values of the electric energy meters are updated.

[0024] In an optional embodiment, the connection state prediction model comprises a feature layer, a time sequence dependence modeling layer, an attention mechanism layer, and a classification output layer,

[0025] The active power, the reactive power, the voltage, the circuit, and the power factor corresponding to the target electric energy meter and the first comparison electric energy meters are input into the feature layer in the form of time sequence data to output a joint feature matrix;

[0026] The time sequence dependence modeling layer performs power consumption behavior modeling on the input joint feature matrix to output modeling features;

[0027] The attention mechanism layer focuses on the abnormal period of the input modeling features to output weighted features;

[0028] The classification output layer performs binary classification based on the input weighted features to output an abnormal probability value as the connection state prediction result.

[0029] In a second aspect, the present application provides a monitoring device for electric energy meter connection, which comprises a monitoring module, used to perform the following steps on a target electric energy meter in a current area according to a preset time interval to determine whether the target electric energy meter is connected abnormally:

[0030] Basic attribute data and historical power consumption data of the target electric energy meter are obtained;

[0031] Based on the completeness of the historical power consumption data of the target electric energy meter, a corresponding comparison electric energy meter set is determined, which is a standard comparison electric energy meter set or a temporary comparison electric energy meter set;

[0032] Based on the basic attribute data of the target electric energy meter, a first comparison electric energy meter is matched in the comparison electric energy meter set;

[0033] Based on the historical power consumption data of the target electric energy meter and the historical power consumption data of each first comparison electric energy meter, a multi-dimension score is performed to obtain a dynamic credibility value of each first comparison electric energy meter;

[0034] The first comparison electric energy meters are ranked based on the dynamic credibility values to determine a second comparison electric energy meter;

[0035] The historical power consumption data corresponding to the target electric energy meter and the second comparison electric energy meter are input into the connection state prediction model to obtain a connection state prediction result, which is used to indicate whether the connection of the target electric energy meter is abnormal.

[0036] In a third aspect, the present application provides an electronic device, comprising a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the electronic device is running, the processor communicates with the memory through the bus, and the processor executes the machine readable instructions to perform the steps of the monitoring method of the electric energy meter connection as described in any of the preceding embodiments.

[0037] In a fourth aspect, the present application provides a computer readable storage medium, the computer readable storage medium stores a computer program, when the computer program is executed by a processor, the steps of the monitoring method of the electric energy meter connection as described in any of the preceding embodiments are performed.

[0038] The present application provides a monitoring method, device, electronic device and storage medium for electric energy meter connection, the method comprising the following steps performed on a target electric energy meter in a current district at a preset time interval to determine whether the connection of the target electric energy meter is abnormal: obtaining basic attribute data and historical power consumption data of the target electric energy meter; determining a corresponding comparison electric energy meter set based on the completeness of the historical power consumption data of the target electric energy meter, the comparison electric energy meter set being a standard comparison electric energy meter set or a temporary comparison electric energy meter set; matching a first comparison electric energy meter in the comparison electric energy meter set based on the basic attribute data of the target electric energy meter; performing multi-dimension scoring based on the historical power consumption data of the target electric energy meter and the historical power consumption data of each first comparison electric energy meter to obtain a dynamic credibility value of each first comparison electric energy meter; ranking the first comparison electric energy meters based on the dynamic credibility values to determine a second comparison electric energy meter; and inputting the historical power consumption data corresponding to the target electric energy meter and the second comparison electric energy meter into a connection state prediction model to obtain a connection state prediction result, which is used to indicate whether the connection of the target electric energy meter is abnormal. Through multi-dimensional, fine-grained and dynamic comparison, the comparison electric energy meter is matched in stages, and then a machine learning model is used to determine whether the connection is abnormal, so that the found comparison electric energy meter is more accurate, and the monitoring result of the remote connection error of the electric energy meter is more accurate. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0040] Figure 1 A flow chart of a monitoring method of an electric energy meter wiring provided by the embodiments of the present application;

[0041] Figure 2 A structural schematic diagram of a power grid system provided by the embodiments of the present application;

[0042] Figure 3 A structural schematic diagram of an electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.

[0044] Embodiment one

[0045] Figure 1 A flow chart of a monitoring method of an electric energy meter wiring provided by the embodiments of the present application. As shown in Figure 1 The monitoring method of the electric energy meter wiring provided by the embodiments of the present application can be executed by the special transformer collection terminal or the intelligent fusion terminal configured corresponding to the transformer area, and specifically can include:

[0046] According to the preset time interval, the following steps are executed on the target electric energy meter in the current transformer area to determine whether the target electric energy meter is wired abnormally.

[0047] S1, the basic attribute data and the historical power consumption data of the target electric energy meter are acquired.

[0048] For each transformer area or each geographical area, a preset time interval can be set to detect each electric energy meter in the range to determine whether the target electric energy meter has wiring errors (including wiring errors caused by human and non-human factors).

[0049] In step S1, the basic attribute data and the historical power consumption data of the target electric energy meter are determined.

[0050] The historical power consumption data here is the power consumption data uploaded by the target electric energy meter in the past period of time, which is usually collected every 15 minutes and can include but is not limited to key electrical parameters such as electric energy (active, reactive), voltage, current, power, power factor, etc.

[0051] The basic attribute data includes, but is not limited to, the wiring mode of the electric energy meter, the communication protocol type, the user type, the device model, and the production batch. These basic attributes can be combined to produce an attribute identifier through field coding. The basic attribute data at least includes the attribute identifier, and the attribute identifier includes a plurality of basic attribute fields, each of which indicates one of the wiring mode, the communication protocol type, the user type, the device model, and the production batch. For example, the attribute identifier can be "DTS188-3P-DLT645-IND-24A01", in which "DTS188" indicates a three-phase electronic electric energy meter, model 188, "3P" indicates three-phase wiring, "DLT645" indicates support for DL / T645-2007 communication protocol, "IND" indicates an industrial user, and "24A01" indicates 2024, A production line, and the first batch.

[0052] S2, based on the completeness of the historical power consumption data of the target electric energy meter, determine the corresponding comparison electric energy meter set, which is a standard comparison electric energy meter set or a temporary comparison electric energy meter set.

[0053] In this embodiment, for all electric energy meters within the range, they can be divided into a standard comparison electric energy meter set and a temporary comparison electric energy meter set based on the completeness of the power consumption data collection and uploading. And the electric energy meter set here will be maintained and updated from time to time.

[0054] For the target electric energy meter, the standard comparison electric energy meter set and the temporary comparison electric energy meter set corresponding to the target electric energy meter can be divided by the following methods:

[0055] For all electric energy meters in the current area that have the same feeder and transformer as the target electric energy meter, determine whether the use duration of the electric energy meter is greater than a preset duration;

[0056] If yes, determine whether the electric energy meter has a fault record, if not, add the electric energy meter to the temporary comparison electric energy meter set;

[0057] If not, determine whether the power consumption data uploaded by the electric energy meter is stable, if so, add the electric energy meter to the temporary comparison electric energy meter set;

[0058] If stable, add the electric energy meter to the standard comparison electric energy meter set, if not stable, add the electric energy meter to the temporary comparison electric energy meter set.

[0059] In this way, for the new area included in the system and the electric energy meter with less data collection, the comparison electric energy meter can be adaptively selected to find a more suitable reference object, and the error caused by subsequent similarity calculation can be avoided. Moreover, at the initial stage of the system going online, the problem of "cold start" with less reference data can be better solved.

[0060] S3, match the first comparison electric energy meter based on the basic attribute data of the target electric energy meter in the comparison electric energy meter set.

[0061] Here, a two-stage comparison electric energy meter screening process can be adopted. In step S3, the basic attributes of the electric meter are first screened based on the attribute identifier, and the candidate electric meter pool can be quickly determined.

[0062] Specifically, the first comparison electric energy meter can be matched in the following way:

[0063] For each electric energy meter in the comparison electric energy meter set, match the attribute identifier of the electric energy meter with the attribute identifier of the target electric energy meter in the basic attribute field.

[0064] If all the basic attribute fields are the same, the electric energy meter is determined to be the first comparison electric energy meter.

[0065] It should be noted that the matching degree of the basic attribute parameter item here can be adjusted. For example, at the initial stage of online in a certain area, the basic attribute parameter item can be relaxed, and only the device model and user type are matched. In the later period, further refined matching of other basic attribute parameter items can be performed.

[0066] S4, based on the historical power consumption data of the target electric energy meter and the historical power consumption data of each first comparison electric energy meter, multi-dimension scoring is performed to obtain the dynamic credibility value of each first comparison electric energy meter.

[0067] S5, sort based on the dynamic credibility value of each first comparison electric energy meter to determine the second comparison electric energy meter.

[0068] In steps S4 and S5, the second stage of screening is performed, and scoring is performed based on the load characteristics and data quality dimensions between the electric energy meters to calculate the dynamic credibility value of each electric energy meter.

[0069] In one possible implementation, the dynamic credibility value of each first comparison electric energy meter can be calculated in the following way:

[0070] Based on the historical power consumption data of the target electric energy meter / first comparison electric energy meter, the daily load curve, monthly load curve, daily load rate interval curve and power fluctuation amount curve are determined. The horizontal coordinate of the curve is time, the sampling step is 15 minutes, and the vertical coordinate is the daily load, monthly load, daily load rate interval and power fluctuation amount corresponding to each time point.

[0071] The load rate interval is the ratio of the average load to the maximum load in the sampling period. The power fluctuation amount is the ratio between the standard deviation and the average value of the user load in the sampling period.

[0072] The distance values corresponding to the daily load curve, the monthly load curve, the daily load rate interval curve and the electricity fluctuation amount curve of the target electric energy meter and the first comparison electric energy are respectively calculated by the dynamic time warping algorithm.

[0073] Based on the calculated plurality of distance values and corresponding weight factors, a dynamic reliability value of the first comparison electric energy meter is calculated by weighting.

[0074] Taking the target electric energy meter A0 and the first comparison electric energy meter A1 as an example, the distance values D1, D2, D3 and D4 corresponding to the daily load curve, the monthly load curve, the daily load rate interval curve and the electricity fluctuation amount curve are respectively calculated by the dynamic time warping algorithm.

[0075] The dynamic reliability value DRS1 corresponding to the first electric energy meter A1 is obtained by weighted summation between D1, D2, D3, D4 and the corresponding weight factors.

[0076] For each first electric energy meter An, the corresponding dynamic reliability value DRSn is calculated, sorted from large to small, and the top 5 to 15 electric energy meters are selected as the second electric energy meter.

[0077] S6, input the historical electricity data corresponding to the target electric energy meter and the second comparison electric energy meter into the connection state prediction model to obtain a connection state prediction result, and the connection state prediction result is used to indicate whether the connection of the target electric energy meter is abnormal.

[0078] In step S6, the connection state prediction model can be built based on AI technology, and the historical electricity data corresponding to the target electric energy meter and the second comparison electric energy meter are analyzed to output the connection state prediction result. Compared with the existing technology which simply relies on electric data parameter calculation and analysis, the classification decision accuracy is higher by using the deep learning model, and the intelligent operation and maintenance level of the power grid is improved.

[0079] The power meter connection monitoring method provided by the application can intelligently select the most reliable comparison object through multi-dimensional scoring and dynamic reliability mechanism, avoid the limitations of single index or fixed reference, make the connection abnormality judgment more accurate and reliable, and effectively cope with complex scenes such as new meter installation, incomplete data or various types of users in the transformer area, and has stronger universality. The whole process does not need manual intervention and can be automatically executed according to the preset period. Through data driving and prediction model, continuous and batch intelligent diagnosis of the connection state of the electric energy meter is realized, and the operation and maintenance efficiency is greatly improved.

[0080] Embodiment two

[0081] In an embodiment of the application, as shown in Figure 2 A power grid system is provided, which includes a plurality of monitoring nodes deployed in different areas. The monitoring nodes can be special transformer acquisition terminals or intelligent fusion terminals.

[0082] Each monitoring node can evaluate and maintain the dynamic credibility of each power meter in the range.

[0083] Specifically, before the step of sorting based on the dynamic credibility values of each first comparison power meter, it further includes:

[0084] Based on the on-site investigation result of the power meter, the dynamic credibility value of the power meter is updated.

[0085] It can be understood that for the target power meter determined by the monitoring node to have an abnormal wiring state, a corresponding work order can be formed for manual on-site inspection. If the on-site investigation result confirms that the power meter wiring is incorrect, the corresponding dynamic credibility value can be increased.

[0086] At the same time, the dynamic division of the comparison power meter set is performed on the power meter.

[0087] In a feasible implementation, the standard comparison power meter set and the temporary comparison power meter set corresponding to the target power meter can be divided by the following method:

[0088] For all power meters in the current area that have the same feeder and transformer as the target power meter, it is determined whether the use duration of the power meter is greater than a preset duration;

[0089] If yes, it is determined whether the power meter has a fault record, and if no, the power meter is added to the temporary comparison power meter set;

[0090] If not, it is determined whether the power consumption data uploaded by the power meter is stable, and if so, the power meter is added to the temporary comparison power meter set;

[0091] If stable, the power meter is added to the standard comparison power meter set, and if not stable, the power meter is added to the temporary comparison power meter set.

[0092] In one specific embodiment, the temporary comparison power meter set is a temporary reference set with low confidence level constructed for the target meter lacking its own historical data.

[0093] The division standard can include (any one of the following conditions can be met):

[0094] The target electric energy meter has a history of valid data storage less than a preset period (e.g., 7 days) since the date of installation and operation. Or the overall operation rate of the area or feeder to which the target electric energy meter belongs is lower than a preset threshold (e.g., 30%), and the system determines that the region is in a cold start state. Or the target meter has no data reported for many days due to communication interruption or other reasons, resulting in a large number of missing data in its historical load curve, which cannot be used for accurate similarity calculation. Or after major power grid reconstruction or events, the system administrator can manually mark the electric energy meters in a specific region as a cold start state to trigger the reconstruction of the reference set.

[0095] In this way, by screening the data quality and health of the electric energy meter, the electric energy meter is divided, and for different target electric energy meters, more matched comparison electric energy meters can be adapted.

[0096] Further, for the electric energy meters in the temporary comparison electric energy meter set, simulation model data can be constructed based on physical rules or typical data across regions as templates to provide virtual reference data for subsequent comparison. In addition, different daily / monthly load curve templates can be configured according to different user types.

[0097] In the early stage of system online and special cases, the accuracy of the system in remotely detecting the connection state can still be maintained.

[0098] In another possible implementation, for the electric energy meters that have been field investigated and have correct connections, they can also be used as seed electric energy meters to evaluate based on the seed electric energy meters and other electric energy meters in the set, while adjusting the dynamic reliability values of each electric energy meter.

[0099] The monitoring node here can also use a combination of triggered and periodic methods to re-evaluate the electric energy meter and divide the set.

[0100] Here, the electric energy meters in the range can be re-evaluated and divided into sets according to the specified period. For example, the accumulation of the electric energy meter historical data volume can be monitored, and when the data volume reaches the threshold, an automatic prompt "re-executable standard screening" is given, and the period can be 7 days.

[0101] In addition, when there is a field investigation result feedback to the system, the dynamic reliability parameter (DRS) of a member in the comparison electric energy meter set falls below the threshold, the topology structure of the power grid where the target electric energy meter is located changes (such as line cutting and modification, transformer capacity increase), the system detects that the load characteristics of the target electric energy meter or the comparison electric energy meter change dramatically and persistently, the electric energy meter evaluation and set division update action can also be triggered.

[0102] In this way, through data accumulation and feedback loops, the cold start situation can be better addressed in the early monitoring stage, and continuous optimization can be ensured.

[0103] Example Three

[0104] In an embodiment of the present application, a wiring state prediction model is provided. The wiring state prediction model comprises a feature layer, a time-dependent modeling layer, an attention mechanism layer, and a classification output layer. The active power, the reactive power, the voltage, the circuit, and the power factor corresponding to the target electric energy meter and the first comparison electric energy meter are input into the feature layer in the form of time series data to output a joint feature matrix. The time-dependent modeling layer models the power consumption behavior of the input joint feature matrix to output a modeling feature. The attention mechanism layer focuses on the abnormal period of the input modeling feature to output a weighted feature. The classification output layer performs binary classification based on the input weighted feature to output an abnormal probability value as the wiring state prediction result.

[0105] In the training process, the historical power consumption data adopts the following measurement data sequence of N time points in the past (for example, 7 days in the past, 15 minutes per point, a total of 7*24*4 = 672 points): active power (P), reactive power (Q), voltage (U), current (I), and power factor (PF).

[0106] For the target electric energy meter (T) and the second comparison electric energy meter (C), each obtains a matrix with a shape of (N, 5), where N is the sequence length and 5 is the feature dimension.

[0107] The final model input concatenates the sequences of the two electric meters in the feature dimension to form a joint input matrix with a shape of (N, 10). The first 5 features are represented as [P_T, Q_T, U_T, I_T, PF_T], and the last 5 features are represented as [P_C, Q_C, U_C, I_C, PF_C] to facilitate the model to learn the relationship between them.

[0108] When constructing the model structure, a dual-flow time series network based on the attention mechanism can be selected.

[0109] The feature layer includes a preprocessing layer and a feature extraction layer. The preprocessing layer inputs the joint sequence with a shape of (N, 10) to perform data standardization and eliminate the dimension effect.

[0110] The feature extraction layer (shared weight) uses a one-dimensional convolutional neural network (1D-CNN) layer to extract local time series patterns and capture short-term and local power consumption features, such as power spikes of device startup, steady-state operation, etc. The convolution kernel can slide on the data of the two electric meters at the same time to preliminarily find the local correlation of the abnormality. Specifically, 64 convolution kernels with a size of 7 can be used, with ReLU activation, followed by a max pooling layer (pooling size = 2) to compress the sequence length.

[0111] The time-dependent modeling layer can use a bidirectional long short-term memory (Bi-LSTM) layer to capture the long-term dependencies and periodicity (e.g., daily cycle, weekly cycle) of the power consumption data. Bi-LSTM can consider both past and future context information, better understanding the power consumption behavior pattern within the entire time window. For example, the normal pattern of "more power consumption during the day and less power consumption during the night", and whether the target meter and the comparison meters are synchronized on this pattern.

[0112] The attention mechanism layer can use self-attention or temporal attention mechanism to let the model focus on those time segments that best expose the abnormal differences. For example, when the comparison meter shows high load, the target meter shows low voltage or abnormal power factor, these time points should be given higher weights.

[0113] The classification output layer integrates the context vectors output by the attention layer through several fully connected layers. Finally, a neuron with a sigmoid activation function outputs a probability value between 0 and 1. The abnormal probability value is output. If the abnormal probability value is greater than a threshold value, it is determined that the connection is abnormal.

[0114] Embodiment Four

[0115] Based on the same inventive concept, the embodiments of the present application also provide a monitoring device for the connection of an electric energy meter. The device includes a monitoring module for performing the following steps on a target electric energy meter in a current area at a predetermined time interval to determine whether the target electric energy meter is connected abnormally:

[0116] Obtaining the basic attribute data and the historical power consumption data of the target electric energy meter;

[0117] Based on the completeness of the historical power consumption data of the target electric energy meter, a corresponding comparison electric energy meter set is determined, which is a standard comparison electric energy meter set or a temporary comparison electric energy meter set;

[0118] Based on the basic attribute data of the target electric energy meter, a first comparison electric energy meter is matched in the comparison electric energy meter set;

[0119] Based on the historical power consumption data of the target electric energy meter and the historical power consumption data of each first comparison electric energy meter, multi-dimensional scoring is performed to obtain a dynamic reliability value of each first comparison electric energy meter;

[0120] Based on the dynamic reliability value of each first comparison electric energy meter, sorting is performed to determine a second comparison electric energy meter;

[0121] The historical power consumption data corresponding to the target electric energy meter and the second comparative electric energy meter are input into the connection state prediction model to obtain a connection state prediction result, which is used to indicate whether the connection of the target electric energy meter is abnormal.

[0122] In a preferred embodiment, the basic attribute data at least includes an attribute identifier, and the attribute identifier includes a plurality of basic attribute fields, each of which indicates one of a connection mode, a communication protocol type, a user type, a device model, and a production batch. The first comparative electric energy meter is matched in the following manner:

[0123] For each electric energy meter in the set of comparative electric energy meters, the attribute identifier of the electric energy meter is matched with the attribute identifier of the target electric energy meter in the basic attribute fields.

[0124] If all the basic attribute fields are the same, the electric energy meter is determined to be the first comparative electric energy meter.

[0125] In a preferred embodiment, the dynamic confidence value of each first comparative electric energy meter is calculated in the following manner:

[0126] Based on the historical power consumption data of the target electric energy meter / first comparative electric energy meter, daily load curves, monthly load curves, daily load rate interval curves, and power fluctuation amount curves are respectively determined.

[0127] The distance values corresponding to the daily load curves, monthly load curves, daily load rate interval curves, and power fluctuation amount curves between the target electric energy meter and the first comparative electric energy meter are respectively calculated by the dynamic time warping algorithm.

[0128] Based on the calculated plurality of distance values and corresponding weight factors, the dynamic confidence value of the first comparative electric energy meter is weightedly calculated.

[0129] In a preferred embodiment, the set of standard comparative electric energy meters corresponding to the target electric energy meter and the set of temporary comparative electric energy meters are divided in the following manner:

[0130] For all electric energy meters in the current area that have the same feeder and transformer as the target electric energy meter, it is determined whether the use duration of the electric energy meter is greater than a preset duration.

[0131] If yes, it is determined whether the electric energy meter has a fault record, and if no, the electric energy meter is added to the set of temporary comparative electric energy meters.

[0132] If not, it is determined whether the power consumption data uploaded by the electric energy meter is stable, and if so, the electric energy meter is added to the set of temporary comparative electric energy meters.

[0133] If stable, the electric energy meter is added to the set of standard comparative electric energy meters, and if not stable, the electric energy meter is added to the set of temporary comparative electric energy meters.

[0134] In a preferred embodiment, the step of determining the corresponding set of comparison electric energy meters based on the completeness of the historical electricity consumption data of the target electric energy meter specifically includes:

[0135] determining whether the use duration of the target electric energy meter is greater than a preset duration, and determining whether the electricity consumption data uploaded by the target electric energy meter is stable;

[0136] If both conditions are met, the standard set of comparison electric energy meters is taken as the set of comparison electric energy meters, otherwise, the temporary set of comparison electric energy meters is taken as the set of comparison electric energy meters.

[0137] In a preferred embodiment, before the step of sorting based on the dynamic credibility values of each first comparison electric energy meter, it further includes:

[0138] updating the dynamic credibility values of the electric energy meters based on the on-site investigation results of the electric energy meters.

[0139] In a preferred embodiment, the connection state prediction model includes a feature layer, a time series dependence modeling layer, an attention mechanism layer, and a classification output layer,

[0140] the active power, the reactive power, the voltage, the circuit, and the power factor corresponding to the target electric energy meter and the first comparison electric energy meter are input into the feature layer in the form of time series data to output a joint feature matrix;

[0141] the time series dependence modeling layer performs electricity consumption behavior modeling on the input joint feature matrix to output modeling features;

[0142] the attention mechanism layer focuses on abnormal time periods for the input modeling features to output weighted features;

[0143] the classification output layer performs binary classification based on the input weighted features to output an abnormal probability value as the connection state prediction result.

[0144] Please refer to Figure 3 , Figure 3 a structural schematic diagram of an electronic device provided by an embodiment of the present application. As shown in Figure 3 , the electronic device 300 includes a processor 310, a memory 320, and a bus 330.

[0145] The memory 320 stores machine-readable instructions executable by the processor 310. When the electronic device 300 is running, the processor 310 and the memory 320 communicate through the bus 330. When the machine-readable instructions are executed by the processor 310, the steps of the monitoring method of the electric energy meter can be performed as shown in the method embodiment described above. For specific implementation manners, please refer to the method embodiment, which will not be described here. Figure 1 ​

[0146] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program. The computer program is run by a processor and can execute one of the above-mentioned method embodiments. Figure 1 The steps of the monitoring method of the electric energy meter wiring in the method embodiment are specifically implemented as described in the method embodiment, which will not be repeated here.

[0147] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-mentioned system, device and unit can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0148] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division, and there can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some communication interface, and can be electrical, mechanical or other forms.

[0149] In addition, the units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. According to actual needs, some or all of the units can be selected to achieve the purpose of the present embodiment.

[0150] Furthermore, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0151] It should be noted that, if the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product in essence or in the form of a part or the technical solutions of the present application that make contributions to the prior art. The computer software product is stored in a storage medium, includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0152] In this article, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations.

[0153] The above only describes the embodiments of the present application and is not used to limit the protection scope of the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method of monitoring the wiring of an electricity meter, characterized in that, The method includes performing the following steps on target electricity meters within the current distribution area at preset time intervals to determine whether the target electricity meters have abnormal wiring: Obtain the basic attribute data and historical electricity consumption data of the target electricity meter; Based on the completeness of the historical electricity consumption data of the target electricity meter, a corresponding set of comparison electricity meters is determined. The set of comparison electricity meters can be a standard set of comparison electricity meters or a temporary set of comparison electricity meters. The step of determining the corresponding set of comparison electricity meters based on the completeness of the historical electricity consumption data of the target electricity meter specifically includes: determining whether the usage time of the target electricity meter is greater than a preset time, and determining whether the electricity consumption data uploaded by the target electricity meter is stable; if both are satisfied, the standard set of comparison electricity meters is used as the set of comparison electricity meters; otherwise, the temporary set of comparison electricity meters is used as the set of comparison electricity meters. Based on the basic attribute data of the target energy meter, the first comparison energy meter is matched in the comparison energy meter set; Multi-dimensional scoring is performed based on the historical electricity consumption data of the target electricity meter and the historical electricity consumption data of each first comparison electricity meter to obtain the dynamic reliability value of each first comparison electricity meter. The second comparison energy meters are determined by sorting them according to the dynamic reliability values ​​of each first comparison energy meter. The historical electricity consumption data corresponding to the target energy meter and the second comparison energy meter are input into the wiring status prediction model to obtain the wiring status prediction result. The wiring status prediction result is used to indicate whether the wiring of the target energy meter is abnormal.

2. The method of claim 1, wherein, The basic attribute data includes at least an attribute identifier, which comprises multiple basic attribute fields. Each basic attribute field indicates one of the following: wiring method, communication protocol type, user type, equipment model, and production batch. The first comparison energy meter is selected by matching in the following way: For each electricity meter in the comparison set, the attribute identifier of the electricity meter is matched with the attribute identifier of the target electricity meter by the basic attribute fields. If all basic attribute fields are the same, then this energy meter is determined to be the first comparison energy meter.

3. The method of claim 2, wherein, The dynamic reliability value of each first comparison energy meter was calculated using the following method: Based on the historical electricity consumption data of the target electricity meter / first comparison electricity meter, the daily load curve, monthly load curve, daily load rate range curve and electricity consumption fluctuation curve are determined respectively. The distance values ​​corresponding to the daily load curve, monthly load curve, daily load rate interval curve and electricity consumption fluctuation curve between the target electricity meter and the first comparison electricity meter are calculated using a dynamic time warping algorithm. Based on the calculated distance values ​​and corresponding weighting factors, the dynamic reliability value of the first comparison energy meter is calculated using a weighted average.

4. The method of claim 3, wherein, The standard comparison energy meter set and the temporary comparison energy meter set corresponding to the target energy are divided in the following way: For all electricity meters in the current distribution area that share the same feeder and transformer as the target electricity meter, determine whether the usage time of the electricity meter is greater than the preset time. If yes, then determine whether the electricity meter has a fault record; if not, then add the electricity meter to the temporary comparison electricity meter set. If it does not exist, determine whether the electricity consumption data uploaded by the electricity meter is stable. If it exists, add the electricity meter to the temporary comparison electricity meter set. If stable, the electric energy meter is added to the standard comparison electric energy meter set, and if unstable, the electric energy meter is added to the temporary comparison electric energy meter set.

5. The method according to claim 1 or 3, characterized in that, Before the step of sorting based on the dynamic credibility values of the first comparison electric energy meters, the method further comprises: Based on the on-site investigation result of the electric energy meter, the dynamic credibility value of the electric energy meter is updated.

6. The method of claim 2, wherein, The wiring state prediction model comprises a feature layer, a time sequence dependence modeling layer, an attention mechanism layer, and a classification output layer, wherein The active power, the reactive power, the voltage, the circuit, and the power factor corresponding to the target electric energy meter and the first comparison electric energy meter are input into the feature layer in the form of time sequence data to output a joint feature matrix; The time sequence dependence modeling layer models the power consumption behavior of the input joint feature matrix to output a modeling feature; The attention mechanism layer focuses on the abnormal period of the input modeling feature to output a weighted feature; The classification output layer performs binary classification based on the input weighted feature to output an abnormal probability value as the wiring state prediction result.

7. A monitoring device for electrical energy meter connections, characterized in that The device comprises a monitoring module configured to perform the following steps on the target electric energy meter in the current area at a preset time interval to determine whether the target electric energy meter is wired abnormally: obtain the basic attribute data and the historical power consumption data of the target electric energy meter; based on the completeness of the historical power consumption data of the target electric energy meter, determine the corresponding comparison electric energy meter set, which is a standard comparison electric energy meter set or a temporary comparison electric energy meter set, the step of determining the corresponding comparison electric energy meter set based on the completeness of the historical power consumption data of the target electric energy meter, specifically comprising: determining whether the use duration of the target electric energy meter is greater than a preset duration, and determining whether the power consumption data uploaded by the target electric energy meter is stable; if both conditions are met, the standard comparison electric energy meter set is used as the comparison electric energy meter set, otherwise, the temporary comparison electric energy meter set is used as the comparison electric energy meter set; based on the basic attribute data of the target electric energy meter, match the first comparison electric energy meter in the comparison electric energy meter set; based on the historical power consumption data of the target electric energy meter and the historical power consumption data of each first comparison electric energy meter, perform multi-dimension scoring to obtain the dynamic credibility value of each first comparison electric energy meter; sort based on the dynamic credibility values of the first comparison electric energy meters to determine the second comparison electric energy meter; input the corresponding historical power consumption data of the target electric energy meter and the second comparison electric energy meter into a wiring state prediction model to obtain a wiring state prediction result, which is used to indicate whether the wiring of the target electric energy meter is abnormal.

8. An electronic device, comprising: comprise: a processor, a memory, and a bus, the memory stores machine readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicate through the bus, the processor executes the machine readable instructions to perform the steps of the electric energy meter wiring monitoring method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, which is executed by the processor to perform the steps of the electric energy meter wiring monitoring method according to any one of claims 1 to 6.

Images