Meter box operation line loss abnormality judgment method and device and computer equipment

Abstract

The application relates to a meter box operation line loss anomaly judgment method and device and computer equipment. The method comprises the following steps: acquiring a historical line loss rate error and a historical line loss rate variance; adaptively correcting the hidden layer weight of an edge intelligent back propagation neural network according to the historical line loss rate error and the historical line loss rate variance to optimize a meter box line loss rate prediction model based on the edge intelligent back propagation neural network; acquiring real-time line loss rate judgment data, acquiring a real-time line loss rate standard value, a real-time line loss rate error and a real-time line loss rate variance according to the real-time line loss rate judgment data and the optimized meter box line loss rate prediction model; and acquiring a meter box operation line loss judgment interval according to the real-time line loss rate standard value, the real-time line loss rate error and the real-time line loss rate variance to judge whether the meter box operation line loss is abnormal. The method can effectively improve the accuracy of meter box operation line loss anomaly judgment.

Description

Technical Field

[0001] This application relates to the field of power grid operation technology, and in particular to a method and device for judging abnormal line loss in meter box operation, as well as a computer device and a readable storage medium. Background Technology

[0002] With the continuous development of power grid operation technology and the continuous improvement of operation and management service levels, electricity metering, as one of the main technical foundations of the power system, has seen its technical level and supervision capabilities continuously improve along with the development of the power grid. Meter boxes, as one of the main metering devices in the power system, play a crucial role in managing electrical energy data. The accuracy and reliability of this data directly determine the effective operation of power services. Managing line loss in meter box operation can effectively achieve electricity monitoring. Among these measures, the judgment of line loss anomalies, as an important component of line loss management, can guide the formulation of loss reduction measures. However, in existing technologies, the accuracy rate of judging line loss anomalies in meter box operation is relatively low. Summary of the Invention

[0003] Therefore, it is necessary to provide a method, device, computer equipment, and readable storage medium for judging meter box operation line loss anomalies, which can improve the accuracy of judging meter box operation line loss anomalies, in order to address the above-mentioned technical problems.

[0004] Firstly, this application provides a method for judging abnormal line loss in meter box operation. The method for judging abnormal line loss in meter box operation includes:

[0005] Obtain historical line loss rate error and historical line loss rate variance;

[0006] The hidden layer weights of the edge-intelligent backpropagation neural network are adaptively adjusted based on the historical line loss rate error and historical line loss rate variance to optimize the meter box line loss rate prediction model based on the edge-intelligent backpropagation neural network.

[0007] Obtain real-time line loss rate determination data, and based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model, obtain the real-time line loss rate standard value, real-time line loss rate error and real-time line loss rate variance;

[0008] The meter box operation line loss judgment range is obtained based on the real-time line loss rate value, real-time line loss rate error, and real-time line loss rate variance to determine whether there is any abnormality in the meter box operation line loss.

[0009] In one embodiment, obtaining the historical line loss rate error and the historical line loss rate variance includes:

[0010] Obtain historical line loss rate determination samples;

[0011] Historical line loss rate error and historical line loss rate variance are obtained based on historical line loss rate judgment samples and meter box line loss rate prediction models.

[0012] In one embodiment, the historical line loss rate determination sample includes historical input samples and historical output samples; the historical line loss rate error and historical line loss rate variance are obtained based on the historical line loss rate determination sample and the meter box line loss rate prediction model, including:

[0013] Historical line loss rate values ​​are obtained based on historical input samples and the meter box line loss rate prediction model;

[0014] The historical line loss rate error is obtained by comparing the historical line loss rate values ​​with historical output samples.

[0015] When the historical line loss rate error exceeds the preset error threshold, the historical line loss rate variance is obtained based on the historical line loss rate calibrated value.

[0016] In one embodiment, the method for determining abnormal line loss in meter box operation further includes:

[0017] When the sum of the real-time line loss rate error and the real-time line loss rate variance does not exceed the preset threshold, a new line loss rate judgment sample is generated based on the real-time line loss rate judgment data, the real-time line loss rate error, and the real-time line loss rate variance, and the new line loss rate judgment sample is sent to the metering automation master station to instruct the updating of the historical line loss rate judgment sample.

[0018] In one embodiment, the hidden layer weights are adaptively adjusted according to the following formula:

[0019] Δx(t)=η(δe+γe′)[(1-μ)g(t)+μg(t-1)]

[0020] In the formula, Δx(t) is the amount of adjustment required to correct the hidden layer weights in the t-th training, η is the learning rate, μ is the heuristic factor, g(t) and g(t-1) are the error gradients calculated in the t-th and t-1-th training respectively, and δ and γ are weight parameters used to measure the importance of historical line loss rate error and historical line loss rate variance respectively.

[0021] In one embodiment, the real-time line loss rate determination data includes real-time input data and real-time output data; obtaining the real-time line loss rate caliber, real-time line loss rate error, and real-time line loss rate variance based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model includes:

[0022] The real-time line loss rate is obtained based on the real-time input data and the optimized meter box line loss rate prediction model;

[0023] The real-time line loss rate error is obtained based on the real-time line loss rate calibrator and the real-time output data, and the real-time line loss rate variance is obtained based on the real-time line loss rate calibrator.

[0024] In one embodiment, the meter box operating line loss determination interval is: in, Let e ​​be the average value of the real-time line loss rate, e be the error of the real-time line loss rate, and e′ be the variance of the real-time line loss rate.

[0025] Secondly, this application also provides a meter box operation line loss anomaly judgment device, the meter box operation line loss anomaly judgment device includes:

[0026] The first parameter acquisition module is used to acquire historical line loss rate error and historical line loss rate variance.

[0027] The model optimization module is used to adaptively adjust the hidden layer weights of the edge intelligent backpropagation neural network based on the historical line loss rate error and historical line loss rate variance to optimize the meter box line loss rate prediction model based on the edge intelligent backpropagation neural network.

[0028] The second parameter acquisition module is used to acquire real-time line loss rate determination data, and to acquire real-time line loss rate calibrated value, real-time line loss rate error and real-time line loss rate variance based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model.

[0029] The anomaly detection module is used to obtain the meter box operation line loss judgment range based on the real-time line loss rate value, real-time line loss rate error and real-time line loss rate variance, so as to determine whether there is an anomaly in the meter box operation line loss.

[0030] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0031] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0032] The aforementioned method and device for judging abnormal line loss in meter boxes, along with computer equipment and a readable storage medium, acquires historical line loss rate error and historical line loss rate variance. Based on these historical line loss rate error and variance, it adaptively adjusts the hidden layer weights of an edge-intelligent backpropagation neural network to optimize the meter box line loss rate prediction model. Furthermore, it acquires real-time line loss rate judgment data. Based on this data and the optimized meter box line loss rate prediction model, it obtains real-time line loss rate values, real-time line loss rate errors, and real-time line loss rate variances. Finally, based on these acquired values, errors, and variances, it obtains the meter box operating line loss judgment interval to determine whether there are any abnormalities in the meter box operating line loss. The aforementioned method for judging abnormal line loss in meter boxes uses real-time line loss rate data for different meter boxes to ultimately obtain the line loss judgment range, allowing the line loss judgment standard to be adjusted according to the differences in meter boxes to adapt to different needs, thereby improving the accuracy of judging abnormal line loss in meter box operation. In addition, the hidden layer weights are adaptively corrected based on historical line loss rate error and historical line loss rate variance, realizing the adaptive adjustment of the hidden layer weight update speed during the backpropagation of the neural network. This can not only improve the convergence speed of the neural network, but also avoid local convergence, thus effectively improving the accuracy of judging abnormal line loss in meter box operation. Attached Figure Description

[0033] Figure 1 This is a flowchart illustrating a method for determining abnormal line loss in meter box operation in one embodiment;

[0034] Figure 2 This is a schematic diagram of the process for obtaining historical line loss rate error and historical line loss rate variance in one embodiment;

[0035] Figure 3 This is a flowchart illustrating the process of obtaining historical line loss rate error and historical line loss rate variance based on historical line loss rate determination samples and meter box line loss rate prediction model in one embodiment.

[0036] Figure 4 This is a flowchart illustrating the process of obtaining the real-time line loss rate value, real-time line loss rate error, and real-time line loss rate variance based on real-time line loss rate determination data and the optimized meter box line loss rate prediction model in one embodiment.

[0037] Figure 5 This is a structural block diagram of a meter box operation line loss anomaly judgment device in one embodiment;

[0038] Figure 6 This is an internal structural diagram of a computer device in one embodiment.

[0039] Explanation of icon numbers:

[0040] Meter box operation line loss anomaly judgment device: 10; First parameter acquisition module: 11; Model optimization module: 12; Second parameter acquisition module: 13; Anomaly judgment module: 14. Detailed Implementation

[0041] To facilitate understanding of the embodiments of this application, a more comprehensive description of the embodiments of this application will be provided below with reference to the accompanying drawings. The drawings illustrate preferred embodiments of the embodiments of this application. However, the embodiments of this application can be implemented in many different forms and are not limited to the embodiments described herein. Rather, these embodiments are provided to make the disclosure of the embodiments of this application more thorough and complete.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which embodiments of this application belong. The terminology used herein in the description of embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the embodiments of this application.

[0043] It is understood that the term "comprising / including" specifies the presence of the stated features, wholes, steps, operations, components, parts, or combinations thereof, but does not preclude the possibility of the presence or addition of one or more other features, wholes, steps, operations, components, parts, or combinations thereof. The terms "first," "second," etc., may be used herein to describe various parameters, but these parameters are not limited by these terms. These terms are used only to distinguish one parameter from another. For example, without departing from the scope of this application, a first parameter acquisition module may be referred to as a second parameter acquisition module, and vice versa.

[0044] Example 1

[0045] By locating anomalies in line loss data and extracting valuable information, timely diagnosis and response to various abnormal operating states affecting power quality can be achieved, thereby reducing line losses and improving energy efficiency. Existing line loss anomaly judgment standards often rely on fixed intervals, failing to differentiate between the number of users, user types, and load levels in different meter boxes. This makes them ill-suited to the complex and ever-changing environment of meter boxes and unable to meet the demand for high-precision line loss anomaly judgment. Secondly, traditional meter box line loss anomaly judgment methods based on backpropagation neural networks have slow convergence speeds and are prone to local convergence, resulting in low neural network prediction accuracy. Finally, the sample library used in neural network training cannot be updated in real time, leading to poor sample data quality and inaccurate meter box line loss rate prediction models constructed based on neural networks. These factors contribute to the low accuracy of existing meter box line loss anomaly judgment technologies, affecting the safe and stable operation of the power grid.

[0046] Compared to other communication technologies, 5G communication technology offers advantages such as high data transmission efficiency, low latency, large system capacity, and low cost, effectively improving the speed and reliability of line-loss data transmission. Edge intelligence technology, a product of the integration of edge computing and artificial intelligence, enables AI algorithms to run on edge devices, effectively improving edge computing efficiency and accuracy. Therefore, to address the problems existing in current technologies, there is an urgent need to propose a method for judging abnormal line loss in meter boxes based on 5G edge intelligence.

[0047] This application provides a method for judging abnormal line loss in meter boxes. Based on 5G edge intelligence, this method can be applied to judging abnormal line loss in meter boxes of power users on the grid side, as well as in the distribution networks of residential areas and buildings, power distribution rooms, and distributed power sources such as photovoltaic and wind turbines. Figure 1 As shown, the method for judging abnormal line loss in meter box operation includes steps 110 to 140.

[0048] Step 110: Obtain the historical line loss rate error and historical line loss rate variance.

[0049] Preferably, such as Figure 2 As shown, step 110 above, obtaining the historical line loss rate error and historical line loss rate variance, includes steps 210 to 220.

[0050] Step 210: Obtain historical line loss rate determination samples. In this embodiment, the intelligent monitoring terminal obtains historical line loss rate determination samples from the metering automation master station sample library via 5G. The historical line loss rate determination samples include historical line loss data, meter readings, user characteristics, load levels, and other data for different meter boxes.

[0051] Step 220: Obtain the historical line loss rate error and historical line loss rate variance based on the historical line loss rate judgment samples and the meter box line loss rate prediction model. In this embodiment, the meter box line loss rate prediction model refers to a neural network used for judging abnormal line loss in meter box operation, which includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer, hidden layer, and output layer are A, B, and C, respectively, and the parameters of the hidden layer are Q(X,Y), where Q is the neural network parameter matrix, X is the hidden layer weight matrix, and Y is the threshold matrix.

[0052] Define the activation function of the edge-intelligent backpropagation neural network suitable for meter box line loss rate prediction as follows:

[0053]

[0054] In the formula, γ is the activation function parameter. This represents the input signal of the a-th input layer neuron.

[0055] Define the hidden layer output function of the edge-intelligent backpropagation neural network suitable for meter box line loss rate prediction as follows:

[0056]

[0057] In the formula, y represents the weights between input layer neurons and hidden layer neurons. b N is the threshold value for the hidden layer. b This is the output of the b-th neuron in the hidden layer.

[0058] Define the output layer function of the edge-intelligent backpropagation neural network suitable for meter box line loss rate prediction as follows:

[0059]

[0060] In the formula, j represents the weights between hidden layer neurons and output layer neurons. c G is the threshold value for the output layer. c This is the output of the c-th neuron.

[0061] Preferably, such as Figure 3 As shown, the historical line loss rate determination sample includes historical input samples and historical output samples. Step 220 above, obtaining the historical line loss rate error and historical line loss rate variance based on the historical line loss rate determination sample and the meter box line loss rate prediction model, includes steps 310 to 330.

[0062] Step 310: Obtain historical line loss rate values ​​based on historical input samples and the meter box line loss rate prediction model. In this embodiment, it should be understood that the historical line loss rate determination samples include historical line loss data, meter readings, user characteristics, load levels, and other data for different meter boxes. Furthermore, the above data can also be classified into historical input samples and historical output samples. By repeatedly iterating through the historical input samples from different meter boxes into the meter box line loss rate prediction model, the meter box line loss rate, i.e., the historical line loss rate values, can be obtained.

[0063] Step 320: Obtain the historical line loss rate error based on the historical line loss rate benchmark and historical output samples. In this embodiment, after obtaining the actual historical output samples and the historical line loss rate benchmark predicted by the meter box line loss rate prediction model based on the historical input samples, that is, after obtaining the actual value and the predicted value, calculate the historical line loss rate error between the two. The historical line loss rate error can be expressed by the following formula:

[0064]

[0065] In the formula, e represents the historical line loss rate error, N represents the number of historical input samples in each training iteration, and g...n denoted as the historical line loss rate of the meter box output by the neural network, and g represents the actual historical output sample of the meter box operation.

[0066] Step 330: When the historical line loss rate error exceeds a preset error threshold, the historical line loss rate variance is obtained based on the historical line loss rate label. In this embodiment, it is determined whether the historical line loss rate error between the actual value and the predicted value meets the requirements. When the requirements are not met, i.e., when the historical line loss rate error exceeds the preset error threshold, the hidden layer weights are updated based on the error gradient. To improve the convergence speed of the edge intelligent backpropagation neural network, this embodiment adopts a weight adaptive update method based on prediction error and variance, i.e., adaptively updating the hidden layer weights based on the historical line loss rate error and the historical line loss rate variance. Therefore, the historical line loss rate variance needs to be obtained based on the historical line loss rate label output by the neural network to achieve adaptive updating of the hidden layer weights. The historical line loss rate error can be expressed by the following formula:

[0067]

[0068] In the formula, e′ represents the historical line loss rate error. This represents the average of the historical line loss rate values ​​for different meter boxes, as output by the neural network.

[0069] Step 120: Adaptively adjust the hidden layer weights of the edge-intelligent backpropagation neural network based on historical line loss rate error and historical line loss rate variance to optimize the meter box line loss rate prediction model based on the edge-intelligent backpropagation neural network. In this embodiment, the hidden layer weights are adaptively updated and adjusted based on historical line loss rate error and historical line loss rate variance. The core idea is that if the current error gradient descent is in the same direction as the previous error gradient descent, the search is accelerated; otherwise, the search is slowed down. Furthermore, when the historical line loss rate error and historical line loss rate variance predicted by the neural network are large, the update speed of the hidden layer weights is increased; conversely, the update speed of the hidden layer weights is decreased. Then, by adaptively updating and adjusting the hidden layer weights, the edge-intelligent backpropagation neural network is trained to optimize the meter box line loss rate prediction model.

[0070] Preferably, the hidden layer weights are adaptively adjusted according to the following formula:

[0071] Δx(t)=η(δe+γe′)[(1-μ)g(t)+μg(t-1)]

[0072] In the formula, Δx(t) is the amount of adjustment required to correct the hidden layer weights in the t-th training, η is the learning rate, μ is the heuristic factor, g(t) and g(t-1) are the error gradients calculated in the t-th and t-1-th training respectively, and δ and γ are weight parameters used to measure the importance of historical line loss rate error and historical line loss rate variance respectively.

[0073] In this embodiment, the hidden layer weight update speed is adaptively adjusted during the backpropagation of the neural network according to the hidden layer weight adaptive update formula. When the historical line loss rate error and historical line loss rate variance are large, the hidden layer weight update speed is increased, and vice versa. This can not only improve the convergence speed of the neural network, but also avoid local convergence, effectively improving the accuracy of meter box operation line loss rate prediction and providing strong support for meter box operation line loss anomaly judgment.

[0074] Step 130: Obtain real-time line loss rate determination data. Based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model, obtain the real-time line loss rate standard value, real-time line loss rate error, and real-time line loss rate variance. In this embodiment, the intelligent monitoring terminal obtains real-time line loss rate determination data for the meter boxes. The real-time line loss rate determination data includes real-time line loss data, meter readings, user characteristics, load levels, and other data for different meter boxes.

[0075] Preferably, such as Figure 4 As shown, the real-time line loss rate determination data includes real-time input data and real-time output data. Step 130 above, which obtains the real-time line loss rate value, real-time line loss rate error and real-time line loss rate variance based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model, includes steps 410 to 420.

[0076] Step 410: Obtain the real-time line loss rate value based on the real-time input data and the optimized meter box line loss rate prediction model. In this embodiment, it should be understood that the real-time line loss rate determination data includes real-time line loss data for different meter boxes, meter readings, user characteristics, load levels, etc. This data can also be categorized into real-time input data and real-time output data. By inputting the real-time input data from different meter boxes into the optimized meter box line loss rate prediction model based on an edge-intelligent backpropagation neural network, the meter box line loss rate, i.e., the real-time line loss rate value, can be obtained.

[0077] Step 420: Obtain the real-time line loss rate error based on the real-time line loss rate benchmark and the real-time output data, and obtain the real-time line loss rate variance based on the real-time line loss rate benchmark. In this embodiment, after obtaining the actual real-time output data and the real-time line loss rate benchmark predicted by the meter box line loss rate prediction model based on the real-time input data, that is, after obtaining the actual value and the predicted value, calculate the real-time line loss rate error between the two, and obtain the real-time line loss rate variance based on the predicted value. The calculation formulas for the real-time line loss rate error and the real-time line loss rate variance can refer to the calculation formulas for the historical line loss rate error and the historical line loss rate variance, respectively, and will not be repeated here.

[0078] Step 140: Obtain the meter box operation line loss judgment range based on the real-time line loss rate value, real-time line loss rate error, and real-time line loss rate variance to determine whether there is any abnormality in the meter box operation line loss.

[0079] Preferably, in one embodiment, the meter box operation line loss determination interval is: in, Let e ​​be the average value of the real-time line loss rate, e be the error of the real-time line loss rate, and e′ be the variance of the real-time line loss rate.

[0080] In this embodiment, when the meter box's line loss rate falls within the corresponding meter box line loss judgment range, it indicates that the meter box's line loss is normal; otherwise, it indicates that the meter box's line loss is abnormal. By using real-time line loss rate judgment data from different meter boxes to ultimately obtain the corresponding meter box line loss judgment range, the problem of fixed line loss rate judgment standards can be solved. This allows for differentiation of differences in the number of users, user types, load levels, etc., among different meter boxes, adapting to the complex and ever-changing environment of the meter boxes and meeting the need for high-precision meter box line loss anomaly judgment.

[0081] Preferably, the method for judging abnormal line loss in meter box operation further includes: when the sum of real-time line loss rate error and real-time line loss rate variance does not exceed a preset threshold, generating a new line loss rate judgment sample based on real-time line loss rate judgment data, real-time line loss rate error and real-time line loss rate variance, and sending the new line loss rate judgment sample to the metering automation master station to instruct the updating of historical line loss rate judgment samples.

[0082] In this embodiment, when the sum of the real-time line loss rate error and the real-time line loss rate variance does not exceed a preset threshold, the new line loss rate judgment sample is transmitted to the metering automation master station via 5G to indicate the updating of the historical line loss rate judgment samples in the sample library. When the preset threshold is exceeded, the samples are abandoned and stored in the sample library, thereby effectively improving the quality of the neural network training samples in the sample library.

[0083] In this embodiment, firstly, the line loss judgment range for meter boxes is finally obtained based on the real-time line loss rate judgment data of different meter boxes, so that the line loss judgment standard can be adjusted according to the differences of meter boxes to adapt to different needs of meter boxes, thereby improving the accuracy of meter box operation line loss anomaly judgment. Secondly, the hidden layer weights are adaptively corrected based on the historical line loss rate error and historical line loss rate variance, realizing the adaptive adjustment of the hidden layer weight update speed during the backpropagation of the neural network. This can not only improve the convergence speed of the neural network, but also avoid local convergence, thereby effectively improving the accuracy of meter box operation line loss anomaly judgment. Finally, whether to update the metering automation master station sample library is determined based on whether the sum of the real-time line loss rate error and the real-time line loss rate variance exceeds a preset threshold. If it exceeds the preset threshold, the sample library update is abandoned to ensure the quality of the neural network training samples in the sample library, thereby effectively improving the accuracy of meter box operation line loss anomaly judgment.

[0084] It should be understood that, although the flowcharts involved in the embodiments described above are... Figures 1-4 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, the flowcharts involved in the embodiments described above... Figures 1-4 At least some of the steps in the process may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but may be executed at different times. The execution order of these steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least some of the steps or stages in other steps.

[0085] Example 2

[0086] Based on the same inventive concept, this application also provides a meter box operation line loss anomaly judgment device for implementing the above-mentioned meter box operation line loss anomaly judgment method. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more meter box operation line loss anomaly judgment device embodiments provided below can be found in the limitations of the meter box operation line loss anomaly judgment method above, and will not be repeated here.

[0087] This application also provides a device for judging abnormal line loss in meter box operation, such as... Figure 5 As shown, the meter box operation line loss anomaly judgment device 10 includes a first parameter acquisition module 11, a model optimization module 12, a second parameter acquisition module 13, and an anomaly judgment module 14. The first parameter acquisition module 11 is used to acquire historical line loss rate error and historical line loss rate variance. The model optimization module 12 is used to adaptively adjust the hidden layer weights of the edge intelligent backpropagation neural network based on the historical line loss rate error and historical line loss rate variance to optimize the meter box line loss rate prediction model based on the edge intelligent backpropagation neural network. The second parameter acquisition module 13 is used to acquire real-time line loss rate judgment data, and obtain the real-time line loss rate scale, real-time line loss rate error, and real-time line loss rate variance based on the real-time line loss rate judgment data and the optimized meter box line loss rate prediction model. The anomaly judgment module 14 is used to obtain the meter box operation line loss judgment interval based on the real-time line loss rate scale, real-time line loss rate error, and real-time line loss rate variance to determine whether there is an anomaly in the meter box operation line loss.

[0088] Preferably, the first parameter acquisition module 11 includes a historical sample acquisition unit and a first parameter acquisition unit. The historical sample acquisition unit is used to acquire historical line loss rate determination samples. The first parameter acquisition unit is used to acquire historical line loss rate error and historical line loss rate variance based on the historical line loss rate determination samples and the meter box line loss rate prediction model.

[0089] Preferably, the historical line loss rate determination sample includes historical input samples and historical output samples. The first parameter acquisition unit is further used to obtain historical line loss rate values ​​based on historical input samples and the meter box line loss rate prediction model; to obtain historical line loss rate errors based on historical line loss rate values ​​and historical output samples; and to obtain historical line loss rate variances based on historical line loss rate values ​​when the historical line loss rate error exceeds a preset error threshold.

[0090] Preferably, the real-time line loss rate determination data includes real-time input data and real-time output data. The second parameter acquisition module 13 includes a prediction data acquisition unit and a second parameter acquisition unit. The prediction data acquisition unit is used to acquire a real-time line loss rate benchmark based on the real-time input data and the optimized meter box line loss rate prediction model. The second parameter acquisition unit is used to acquire the real-time line loss rate error based on the real-time line loss rate benchmark and the real-time output data, and to acquire the real-time line loss rate variance based on the real-time line loss rate benchmark.

[0091] Preferably, the above-mentioned meter box operation line loss anomaly judgment device 10 further includes a new sample sending module, which is used to generate a new line loss rate judgment sample based on the real-time line loss rate judgment data, the real-time line loss rate error and the real-time line loss rate variance when the sum of the real-time line loss rate error and the real-time line loss rate variance does not exceed a preset threshold, and send the new line loss rate judgment sample to the metering automation master station to indicate the update of the historical line loss rate judgment sample.

[0092] Each module in the aforementioned meter box operation line loss anomaly judgment device 10 can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0093] Example 3

[0094] This application also provides a computer device, such as... Figure 6 As shown, it includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the above-mentioned method for judging abnormal line loss in meter box operation.

[0095] Those skilled in the art will understand that Figure 6The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0096] Example 4

[0097] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method for judging abnormal line loss in meter box operation.

[0098] Example 5

[0099] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for judging abnormal line loss in meter box operation.

[0100] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0101] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0102] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for judging abnormal line loss in meter box operation, characterized in that, The method includes: Obtain historical line loss rate error and historical line loss rate variance; The hidden layer weights of the edge intelligent backpropagation neural network are adaptively adjusted based on the historical line loss rate error and the historical line loss rate variance to optimize the meter box line loss rate prediction model based on the edge intelligent backpropagation neural network. Obtain real-time line loss rate determination data, and obtain real-time line loss rate calibrated value, real-time line loss rate error and real-time line loss rate variance based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model; The meter box operation line loss determination interval is obtained based on the real-time line loss rate value, the real-time line loss rate error, and the real-time line loss rate variance, in order to determine whether there is any abnormality in the meter box operation line loss; The hidden layer weights are adaptively adjusted according to the following formula: In the formula, For the first The next training iteration requires adjusting the amount of weights in the hidden layer. For learning rate, As a heuristic factor, and The first Second and third The error gradient calculated during the training. and These are weighting parameters, used to measure the importance of the historical line loss rate error and the historical line loss rate variance, respectively. The real-time line loss rate error is... Let V be the variance of the real-time line loss rate.

2. The method according to claim 1, characterized in that, The acquisition of historical line loss rate error and historical line loss rate variance includes: Obtain historical line loss rate determination samples; The historical line loss rate error and the historical line loss rate variance are obtained based on the historical line loss rate judgment sample and the meter box line loss rate prediction model.

3. The method according to claim 2, characterized in that, The historical line loss rate determination sample includes historical input samples and historical output samples; the step of obtaining the historical line loss rate error and the historical line loss rate variance based on the historical line loss rate determination sample and the meter box line loss rate prediction model includes: Historical line loss rate values ​​are obtained based on the historical input samples and the meter box line loss rate prediction model. The historical line loss rate error is obtained by comparing the historical line loss rate calibrated value with the historical output sample. When the historical line loss rate error exceeds a preset error threshold, the historical line loss rate variance is obtained based on the historical line loss rate value.

4. The method according to claim 2, characterized in that, The method further includes: When the sum of the real-time line loss rate error and the real-time line loss rate variance does not exceed a preset threshold, a new line loss rate judgment sample is generated based on the real-time line loss rate judgment data, the real-time line loss rate error, and the real-time line loss rate variance, and the new line loss rate judgment sample is sent to the metering automation master station to instruct the updating of the historical line loss rate judgment sample.

5. The method according to claim 1, characterized in that, The historical line loss rate determination sample includes historical line loss data, meter readings, user characteristics, and load levels for different meter boxes; the real-time line loss rate determination data includes real-time line loss data, meter readings, user characteristics, and load levels for different meter boxes.

6. The method according to claim 1, characterized in that, The real-time line loss rate determination data includes real-time input data and real-time output data; the step of obtaining the real-time line loss rate calibrator, real-time line loss rate error, and real-time line loss rate variance based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model includes: The real-time line loss rate value is obtained based on the real-time input data and the optimized meter box line loss rate prediction model; The real-time line loss rate error is obtained based on the real-time line loss rate calibrator and the real-time output data, and the real-time line loss rate variance is obtained based on the real-time line loss rate calibrator.

7. The method according to claim 1, characterized in that, The meter box operating line loss determination range is: ,in, This is the average value of the real-time line loss rate calibrator.

8. A device for judging abnormal line loss in meter box operation, characterized in that, The device includes: The first parameter acquisition module is used to acquire historical line loss rate error and historical line loss rate variance. The model optimization module is used to adaptively correct the hidden layer weights of the edge intelligent backpropagation neural network based on the historical line loss rate error and the historical line loss rate variance to optimize the meter box line loss rate prediction model based on the edge intelligent backpropagation neural network. The second parameter acquisition module is used to acquire real-time line loss rate determination data, and to acquire real-time line loss rate calibrator, real-time line loss rate error and real-time line loss rate variance based on the real-time line loss rate determination data and the optimized meter box line loss rate prediction model. The anomaly detection module is used to obtain the meter box operation line loss judgment interval based on the real-time line loss rate label, the real-time line loss rate error and the real-time line loss rate variance, so as to determine whether there is an anomaly in the meter box operation line loss; The hidden layer weights are adaptively adjusted according to the following formula: In the formula, For the first The next training iteration requires adjusting the amount of weights in the hidden layer. For learning rate, As a heuristic factor, and The first Second and third The error gradient calculated during the training. and These are weighting parameters, used to measure the importance of the historical line loss rate error and the historical line loss rate variance, respectively. The real-time line loss rate error is... Let V be the variance of the real-time line loss rate.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images