Power consumer side line loss monitoring and diagnosis method and system, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a power consumer side line loss monitoring and diagnosis method and system, electronic equipment and a storage medium, and belongs to the field of line loss monitoring and diagnosis. The method comprises the following steps: collecting power utilization active and reactive power data of a general meter and sub-meters; using the total table power prediction model to obtain predicted total table power data, comparing the predicted total table power data with actually measured total table power data, and performing abnormality judgment in combination with a preset monitoring rule; and when abnormality is detected, different correction modes are adopted to carry out correction processing on the sub-table data, prediction residual errors under different correction modes are obtained by optimizing and minimizing prediction errors, and the correction mode with the residual errors from small to large is selected as a diagnosis result. According to the method, the overall power consumption condition of the power system is estimated under the condition that the total meter power prediction model is established without depending on real-time branch head and tail end electric meter readings, and when abnormality is detected, different correction modes are adopted to perform positioning processing on branch meter power data, so that the positioning accuracy and the response rapidness are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of line loss monitoring and diagnosis, and in particular to a method, system, electronic equipment and storage medium for monitoring and diagnosing line loss on the power user side. Background Art

[0002] Calculation based on a fixed ratio or electricity consumption ratio occurs after actual line losses are calculated. However, abnormal line losses are only detected when fluctuations in the percentage of the settlement period (i.e., the ratio to total electricity consumption) exceed a certain threshold, triggering a comprehensive investigation. Firstly, the long cycle makes it difficult to detect abnormalities in a timely manner; secondly, it is difficult to quickly diagnose the abnormal location, requiring comprehensive manual investigation.

[0003] Traditional power flow calculation methods solve a set of equations to obtain line loss rates, requiring a large number of test quantities (test instruments) and fixed quantities (line topology, electrical parameters, etc.). This increases the cost of instrument installation and maintenance and parameter testing. It is also difficult to update line topology and parameters in a timely manner when they change over time or due to human influence, making the solution difficult. Summary of the Invention

[0004] The purpose of the embodiments of the present invention is to provide a method, system, electronic device and storage medium for monitoring and diagnosing power user-side line losses, which are used to fully or at least partially solve the technical problems existing in the above-mentioned prior art.

[0005] To achieve the above objectives, an embodiment of the present invention provides a method for monitoring and diagnosing power user-side line losses, comprising: Collect active power and reactive power data of the main meter and sub-meters in normal operation over a period of time; The pre-built total meter power prediction model is used to predict the collected active power and reactive power data of the sub-meter electricity consumption to obtain the predicted total meter power data. The predicted total meter power data is then compared with the measured total meter power data, and abnormality judgment is performed in combination with the preset monitoring rules; When an anomaly is detected, different correction methods are used to correct the sub-meter power data, and a new prediction is made based on the corrected data. By optimizing and minimizing the prediction error, the prediction residuals under different correction methods are obtained. The correction methods with the smallest to largest residuals are selected as the diagnosis results, the diagnosis results are output, and an alarm is processed according to the set threshold conditions.

[0006] Optionally, the process of building a total meter power prediction model includes: The MLP neural network model is used as the basic framework of the total meter power prediction model. The active power and reactive power data of the sub-meter electricity consumption are used as input, and the active power and reactive power data of the total meter electricity consumption are used as output. A total meter power prediction model is established. The MLP neural network model includes an input layer, a hidden layer, and an output layer.

[0007] Optionally, the predicted total meter power data is compared with the measured total meter power data to obtain the prediction residual, and the predicted normal fluctuation range parameters are determined based on the prediction residual.

[0008] Optionally, the correction processing steps include: Establish a correction equation to correct the power data of the terminal meter, and then re-input the corrected power data and the non-corrected data into the total meter power prediction model for prediction; The least square method is used to determine the best fitting coefficient of the correction equation so as to minimize the correction prediction residual; Compare the prediction residuals after correction at different terminals, determine the order of the line segments where abnormalities occur, and conduct investigations in the order where the abnormalities occur.

[0009] Optionally, the correction equation is in the form of a linear combination equation, including a linear combination of active power and reactive power, and data correction is achieved by adjusting coefficients.

[0010] Optionally, the least squares method is used to determine the best fitting coefficients of the correction equation to minimize the correction prediction residual, including: Step 1: Choose a coding for problem solving; Step 2: Randomly generate N feasible solutions, that is, generate the first generation of chromosomes. The parameter N is the population size of chromosomes. Step 3: Then use the fitness function to calculate the fitness of each chromosome, and determine whether the chromosome group meets the termination condition. If so, go to step 8. Step 4: Calculate the probability of each chromosome being selected in the next evolution based on the fitness level and calculation formula; Step 5: Perform replication. From the current chromosome group consisting of N chromosomes, select chromosomes according to the probability of each chromosome being selected calculated in step 4, and select M chromosomes as part of the next generation chromosome group, where M <N; Step 6: Perform a crossover operation within this group of M chromosomes with the mating probability to generate NM chromosomes. Then perform a mutation operation on the NM chromosomes generated after the crossover. The mutated NM chromosomes are merged with the M chromosomes generated by replication to form a new chromosome group. Step 7: Return to step 3 and check whether the newly generated chromosome population meets the termination condition. If it does, proceed to step 8. If not, proceed to the next round of iteration. Step 8: Terminate the calculation and output the optimal result.

[0011] On the other hand, the present invention also provides a power user-side line loss monitoring and diagnosis system, comprising: The collection unit is used to collect the active power and reactive power data of the main meter and sub-meters that are operating normally over a period of time; The prediction unit is used to predict the collected active power and reactive power data of the sub-meter electricity consumption using a pre-built total meter power prediction model to obtain the predicted total meter power data, compare the predicted total meter power data with the measured total meter power data, and make an abnormality judgment based on the preset monitoring rules; The diagnostic unit is used to correct the sub-meter power data using different correction methods when an abnormality is detected, and determine the coefficient of the correction method based on minimizing the prediction error of the corrected data, and compare the prediction residuals under different correction methods, select the correction methods with the smallest residuals to the largest as the diagnosis result, output the diagnosis result, and perform alarm processing according to the set threshold conditions.

[0012] Optionally, the diagnostic unit is specifically configured to: Establish a correction equation to correct the power data of the terminal meter, and then re-input the corrected power data and the non-corrected data into the total meter power prediction model for prediction; The least square method is used to determine the best fitting coefficient of the correction equation so as to minimize the correction prediction residual; Compare the prediction residuals after correction at different terminals, determine the order of the line segments where abnormalities occur, and conduct investigations in the order where the abnormalities occur.

[0013] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for monitoring and diagnosing line losses on the power user side when executing the program.

[0014] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which implements the steps of the above-mentioned method for monitoring and diagnosing power user-side line losses when executed by a processor.

[0015] Through the above technical solution, the basic configuration of the electric meter with a main meter plus branch meters does not require the calculation of line loss rate, which reduces costs. In addition, by establishing a main meter power prediction model, the overall power consumption of the power system can be estimated without relying on real-time main meter readings. When an anomaly is detected, different correction methods can be used to correct the sub-meter power data, which has high flexibility, enabling it to deal with different types of anomalies and find the most appropriate solution.

[0016] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following detailed description, they are used to explain the embodiments of the present invention, but do not constitute a limitation of the embodiments of the present invention. In the accompanying drawings: Figure 1 This is a flowchart of an implementation method for monitoring and diagnosing power user-side line losses provided by an embodiment of the present invention; Figure 2 This is a flowchart of a correction process implementation provided by an embodiment of the present invention; Figure 3 This is a flowchart of an implementation of determining the best fitting coefficient of a correction equation provided by an embodiment of the present invention; Figure 4 This is a detailed implementation flow chart of a method for monitoring and diagnosing power user-side line losses provided by an embodiment of the present invention; Figure 5 This is a power topology diagram provided by an embodiment of the present invention; Figure 6 This is a structural diagram of a power user-side line loss monitoring and diagnosis system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following describes the specific implementation of the embodiment of the present invention in detail with reference to the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the embodiment of the present invention and is not used to limit the embodiment of the present invention.

[0019] See Figure 1 FIG. 1 is a flowchart of an implementation method for monitoring and diagnosing power user-side line losses provided by an embodiment of the present invention, including the following steps: Step 100: Collect active power and reactive power data of the main meter and sub-meters that are operating normally over a period of time.

[0020] Step 101: Use the pre-built total meter power prediction model to predict the collected sub-meter electricity active power and reactive power data to obtain the predicted total meter power data, compare the predicted total meter power data with the measured total meter power data, and make an abnormality judgment based on the preset monitoring rules.

[0021] In some embodiments, the process of constructing the total meter power prediction model includes: using the MLP neural network model as the basic architecture of the total meter power prediction model, using the active power and reactive power data of the sub-meter electricity consumption as input, and the active power and reactive power data of the total meter electricity consumption as output, to establish the total meter power prediction model, wherein the MLP neural network model includes an input layer, a hidden layer, and an output layer.

[0022] In some embodiments, when executing step 101, the predicted total meter power data is compared with the measured total meter power data to obtain a prediction residual, and a predicted normal fluctuation range parameter is determined based on the prediction residual.

[0023] Specifically, the preset monitoring rules used when executing step 101 for abnormal judgment are multiple judgment criteria, including but not limited to: a single point exceeds area A, multiple consecutive points exceed area B or area C, multiple consecutive points increase or decrease, and multiple consecutive points are distributed on one side of the center line but do not fall within area C, among which area A, area B, and area C are determined by normal fluctuation range parameters.

[0024] In some embodiments, when standard deviation is used, Region C is [-σ, σ], Region B is [-2σ, -σ] and [σ, 2σ], and Region A is [-3σ, -2σ] and [2σ, 3σ]. σ is substituted for σp and σQ, respectively. When quantiles are used, Region C is [q1-0.5IQR, q3+0.5IQR], Region B is [q1-IQR, q1-0.5IQR] and [q3+0.5IQR, q3+IQR], and Region A is [q1-1.5IQR, q1-IQR] and [q3+IQR, q3+1.5IQR]. IQR = q3–q1, with the fractional digits of P and Q substituted for the respective values.

[0025] Step 102: When an abnormality is detected, different correction methods are used to correct the sub-meter power data, and a new prediction is made based on the corrected data. By minimizing the prediction error, the prediction residuals under different correction methods are obtained. The correction methods with the residuals from small to large are selected as the diagnosis results, the diagnosis results are output, and an alarm is processed according to the set threshold conditions.

[0026] In some embodiments, when performing step 102, refer to Figure 2 FIG. 1 is a flowchart of a correction process implementation provided by an embodiment of the present invention, including the following execution steps: S1020: Establish a correction equation to correct the power data of the terminal electricity meter, and re-input the corrected power data and the non-corrected data into the total meter power prediction model for prediction.

[0027] The correction equation is in the form of a linear combination equation, including a linear combination of active power and reactive power, and data correction is achieved by adjusting coefficients.

[0028] S1021: Use the least squares method to determine the best fitting coefficient of the correction equation to minimize the correction prediction residual.

[0029] S1022: Compare the prediction residuals after correction at different terminals, determine the line segment where the abnormality occurs, and conduct an investigation in the order of the line segments where the abnormality occurs.

[0030] In some embodiments, see Figure 3 As shown, when executing step S1021, the following steps may be specifically performed: Step 1: Choose a coding for problem solving; Step 2: Randomly generate N feasible solutions, that is, generate the first generation of chromosomes. The parameter N is the population size of chromosomes. Step 3: Then use the fitness function to calculate the fitness of each chromosome and determine whether the chromosome group meets the termination condition. If so, go to step 8; otherwise, go to step 4. Step 4: Calculate the probability of each chromosome being selected in the next evolution based on the fitness level and calculation formula; Step 5: Perform replication. From the current chromosome group consisting of N chromosomes, select chromosomes according to the probability of each chromosome being selected calculated in step 4, and select M chromosomes as part of the next generation chromosome group, where M <N; Step 6: Perform a crossover operation within this group of M chromosomes with the mating probability to generate NM chromosomes. Then perform a mutation operation on the NM chromosomes generated after the crossover. The mutated NM chromosomes are merged with the M chromosomes generated by replication to form a new chromosome group. Step 7: Return to step 3 and check whether the newly generated chromosome population meets the termination condition. If it does, proceed to step 8. If not, proceed to the next round of iteration. Step 8: Terminate the calculation and output the optimal result.

[0031] The calculation formula in step 4 is: the probability of chromosome i being selected = the fitness of chromosome i / the sum of the fitness of all chromosomes.

[0032] See Figure 4 FIG. 1 is a detailed implementation flow chart of a method for monitoring and diagnosing power user-side line loss provided by an embodiment of the present invention, including the following steps: S400: Collecting active power and reactive power data of the main meter and sub-meters that are operating normally over a period of time.

[0033] S401: Using the MLP neural network model as the basic architecture of the total meter power prediction model, using the active power and reactive power data of the sub-meter electricity consumption as input and the active power and reactive power data of the total meter electricity consumption as output, a total meter power prediction model is established, wherein the MLP neural network model includes an input layer, a hidden layer, and an output layer.

[0034] S402: Use the pre-built total meter power prediction model to predict the collected sub-meter electricity active power and reactive power data to obtain the predicted total meter power data, and compare the predicted total meter power data with the measured total meter power data, and make an abnormality judgment based on the preset monitoring rules.

[0035] S403: When an abnormality is detected, a correction equation is established to correct the power data of the terminal electricity meter, and the corrected power data is combined with the non-corrected data and re-input into the total meter power prediction model for prediction.

[0036] S404: Randomly give values ​​of fitting coefficients of multiple sets of correction equations, and substitute them into the multiple sets of correction equations to obtain new sub-meter power data and errors.

[0037] S405: Selecting a correction equation with a small error from the multiple sets of correction equations, and fine-tuning its fitting coefficient to obtain a new error.

[0038] S406: Selecting a correction equation with a smaller error among the new errors, continuing fine-tuning and selecting until a stop condition is reached, and selecting the fitting coefficient corresponding to the result with the smallest error as the best fitting coefficient obtained by optimization.

[0039] S407: Using the least squares method to determine the best fitting coefficient of the correction equation so as to minimize the correction prediction residual.

[0040] S408: Select the correction method of the residual from small to large as the diagnosis result, output the diagnosis result, and perform alarm processing according to the set threshold condition.

[0041] In one embodiment, see Figure 5 As shown in FIG. 1 , a power topology diagram is provided in an embodiment of the present invention. Figure 5 The power topology diagram shown in FIG is used as an example to further explain in detail a method for monitoring and diagnosing line loss on the power user side provided by the present invention. Figure 5 In the figure, 0 is the main table, 1 to 5 are the end-user sub-tables, and A, B, C, D, E, F, and G are line nodes respectively.

[0042] 1. Collect periodic (e.g., 5-minute) active and reactive power records for the master and sub-meters for normal operation over a longer period (e.g., one year, three months, etc.): P0-P5, Q1-Q5. Perform data preprocessing, such as deleting records with missing measurements (null values) and outliers (e.g., those exceeding the meter range).

[0043] 2. Using P1-P5 and Q1-Q5 as inputs and P0 and Q0 as outputs, an MLP neural network model is established and trained using collected data to obtain a total meter power prediction model. The difference between the predicted and actual values ​​for each record, known as the residual, is also obtained. Statistical processing of the residuals yields the standard deviations σP and σQ of the two outputs, or the first quartiles q1P and q1Q and the third quartiles q3P and q3Q, etc. Monitoring indicators for the predicted power difference are then derived. When using standard deviations, the values ​​for Region C are [-σ, σ]; for Region B, [-2σ, -σ] and [σ, 2σ]; and for Region A, [-3σ, -2σ] and [2σ, 3σ]. Substitute σ for σP and σQ, respectively. When using quantiles, Area C is [q1-0.5IQR, q3+0.5IQR], Area B is [q1-IQR, q1-0.5IQR] and [q3+0.5IQR, q3+IQR], and Area A is [q1-1.5IQR, q1-IQR] and [q3+IQR, q3+1.5IQR]. Where IQR = q3 – q1, substitute the fractional digits of P and Q respectively.

[0044] 3. Use the obtained model to monitor real-time periodic data collection. That is, use the model input to collect P1-P5 and Q1-Q5 in real time, obtain the predicted P'0 and Q'0, and compare them with the real-time collected P0 and Q0 to obtain their differences. After collecting a set of data (such as half an hour, 6 period measurement), calculate the average value. Use the judgment criteria to judge the average of multiple groups, for example: Criterion 1: 1 point is outside zone A (greater than Ucl or less than Lcl); Criterion 2: 2 out of 3 consecutive points are outside of Area B on one side; Criterion 3: 4 out of 5 consecutive points are outside Zone C; Criterion 4: Six consecutive points of increase or decrease.

[0045] Considering that abnormalities are generally persistent, in order to reduce false alarms, a threshold for the number of consecutive triggers can be set (such as more than 5 times), and an abnormal alarm will be issued only when it is exceeded.

[0046] 4. When there is an abnormal alarm, perform diagnosis and processing.

[0047] 4.1. Correct the terminal meter data sequentially. For example, when correcting terminal 1, establish correction equations for P1 and Q1: P'1 = aP1 + bQ1 + c; Q'1 = dP1 + eQ1 + f, where a, b, c, d, e, and f are the coefficients to be fitted. Combine the corrected data from the abnormal group with other uncorrected data, such as P2, Q2, P3, Q3, and the neural network model, to obtain the new total meter output.

[0048] 4.2. Use genetic algorithm for optimization (1. Randomly give multiple groups of ab values, substitute them to obtain new P'1Q'1 instead of P1Q1, and use the method in step 2 to obtain a new model and the error of the model output (the sum of the squares of the difference between the output value and the actual value); 2. Select several groups with smaller errors from multiple groups, and randomly (genetically, mutate) fine-tune ab (taking coefficient ab as an example), and re-obtain new errors; 3. Select several groups with smaller errors from the new results and previous results, and continue fine-tuning and selecting until the stopping condition is reached (number limit or no improvement in the result); 4. Select the ab corresponding to the result with the smallest error, which is the best fitting coefficient obtained by optimization), and obtain the best fitting formula (coefficient) to minimize the error between the output value of the corrected input neural network model and the actual measurement.

[0049] 4.3. Compare the minimum errors after correction at different terminals to prioritize the terminal to be detected. For example, if the error after correction at terminal 3 is the smallest, it is more likely that the branch where terminal 3 is located (D-E segment) has an anomaly.

[0050] 4.4. Based on the topology, make corrections in sequence along the selected priority terminal direction. For example, the active power of the CD segment is (P2+P3) and the reactive power is (Q2+Q3); the active power of the BC segment is (P2+P3+P4+P5) and the reactive power is (Q1+Q2+Q3+Q4), and so on. Establish the correction equation similar to 4.1. If the CD segment includes terminals 2 and 3, make the correction together: P'2 = aP2+bQ2+c; Q'2 = dP2+eQ2+f; P'3 = aP3+bQ3+c; Q'3 = dP3+eQ3+f; Similarly, combine the uncorrected data (except for terminals 2 and 3) with the neural grid to establish an error calculation function. 4.5. Similar to 4.2, obtain the prediction residuals for the optimal fit coefficients. 4.6. Similar to 4.3, sort and determine the line segments most likely to experience anomalies.

[0051] 5. Check the abnormal line sections according to the results in 4.6.

[0052] 6. Summarize and accumulate experience, namely the relationship between different abnormal locations, causes, and fitting coefficients. For example, an increase in the corresponding resistance will cause the active fitting coefficient to increase, and an increase in reactance will cause the reactive fitting coefficient to increase. This can improve the pertinence and efficiency of the troubleshooting in step 5.

[0053] See Figure 6 FIG. 1 is a schematic diagram of a system for monitoring and diagnosing power user-side line losses according to an embodiment of the present invention, comprising: The collection unit 60 is used to collect the active power and reactive power data of the main meter and sub-meters that are operating normally over a period of time; The prediction unit 61 is used to use the pre-built total meter power prediction model to predict the collected sub-meter power active power and reactive power data to obtain predicted total meter power data, compare the predicted total meter power data with the measured total meter power data, and perform abnormality judgment based on preset monitoring rules; The diagnostic unit 62 is used to correct the sub-meter power data using different correction methods when an abnormality is detected, and to determine the coefficient of the correction method based on minimizing the prediction error of the corrected data, and to compare the prediction residuals under different correction methods, select the correction methods with the smallest residuals to the largest as the diagnostic results, output the diagnostic results, and perform alarm processing according to the set threshold conditions.

[0054] In some embodiments, the diagnostic unit 62 is specifically configured to: Establish a correction equation to correct the power data of the terminal meter, and then re-input the corrected power data and the non-corrected data into the total meter power prediction model for prediction; The least square method is used to determine the best fitting coefficient of the correction equation so as to minimize the correction prediction residual; Compare the prediction residuals after correction at different terminals, determine the order of the line segments where abnormalities occur, and conduct investigations in the order where the abnormalities occur.

[0055] Through the above technical solution, the basic configuration of the electric meter with a main meter plus branch meters does not require the calculation of line loss rate, which reduces costs. Moreover, by establishing a main meter power prediction model, the system can estimate the overall power consumption of the power system without relying on real-time main meter readings. When an anomaly is detected, the system can use different correction methods to correct the sub-meter power data. It has high flexibility, enabling the system to deal with different types of anomalies and find the most appropriate solution.

[0056] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for monitoring and diagnosing line losses on the power user side when executing the program.

[0057] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which implements the steps of the above-mentioned method for monitoring and diagnosing power user-side line losses when executed by a processor.

[0058] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0059] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0060] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0061] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0062] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0063] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0064] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can be implemented using any method or technology to store information. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change RAM (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media such as modulated data signals and carrier waves.

[0065] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0066] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for monitoring and diagnosing line loss on the power user side, characterized in that: include: Collect active power and reactive power data of the total meter and sub-meters that are operating normally over a period of time; The pre-built total meter power prediction model is used to predict the collected sub-meter power consumption active power and reactive power data to obtain the predicted total meter power data, and the predicted total meter power data is compared with the measured total meter power data, and abnormal judgment is made in combination with the preset monitoring rules; When an abnormality is detected, different correction methods are used to correct the sub-meter power data, and a new prediction is made based on the corrected data. By optimizing and minimizing the prediction error, the prediction residuals under different correction methods are obtained. The correction methods with the residuals from small to large are selected as the diagnosis results, the diagnosis results are output, and alarm processing is performed according to the set threshold conditions.

2. The method for monitoring and diagnosing line loss at the power user side according to claim 1, characterized in that: The construction process of the total meter power prediction model includes: The MLP neural network model is used as the basic framework of the total meter power prediction model. The active power and reactive power data of the sub-meter electricity consumption are used as input, and the active power and reactive power data of the total meter electricity consumption are used as output. The total meter power prediction model is established, wherein the MLP neural network model includes an input layer, a hidden layer and an output layer.

3. The method for monitoring and diagnosing line loss at the power user side according to claim 1, characterized in that: The predicted total meter power data is compared with the measured total meter power data to obtain the predicted residual, and the predicted normal fluctuation range parameters are determined based on the predicted residual.

4. The method for monitoring and diagnosing line loss at the power user side according to claim 1, characterized in that: The correction process steps include: Establish a correction equation to correct the power data of the terminal meter, and re-input the corrected power data and the non-corrected data into the total meter power prediction model for prediction; The least square method is used to determine the best fitting coefficient of the correction equation to minimize the prediction residual after correction; Compare the prediction residuals after correction at different ends, determine the order of the line segments with abnormalities, and conduct inspections in the order of the line segments with abnormalities.

5. The method for monitoring and diagnosing line loss at the power user side according to claim 4, characterized in that: The correction equation is in the form of a linear combination equation, including a linear combination of active power and reactive power, and data correction is achieved by adjusting coefficients.

6. The method for monitoring and diagnosing line loss at the power user side according to claim 4, characterized in that: The least squares method is used to determine the best fitting coefficients of the correction equation to minimize the corrected prediction residuals, including: Step 1: Choose a coding for problem solving; Step 2: Randomly generate N feasible solutions, that is, generate the first generation of chromosomes. The parameter N is the population size of chromosomes. Step 3: Then use the fitness function to calculate the fitness of each chromosome, and determine whether the chromosome group meets the termination condition at this time. If it does, go to step 8; Step 4: Calculate the probability of each chromosome being selected in the next evolution based on the fitness level and calculation formula; Step 5: Perform replication. From the current chromosome group consisting of N chromosomes, select chromosomes according to the probability of each chromosome being selected calculated in step 4, and select M chromosomes as part of the next generation chromosome group, where M <N; Step 6: Perform a crossover operation with the mating probability in this group of M chromosomes to generate NM chromosomes, then perform a mutation operation on the NM chromosomes generated after the crossover, and merge the mutated NM chromosomes with the M chromosomes generated by replication to generate a new chromosome group; Step 7: Return to step 3 and check whether the newly generated chromosome population meets the termination condition. If it does, proceed to step 8. If not, proceed to the next round of iteration. Step 8: Terminate the calculation and output the optimal result.

7. A power user side line loss monitoring and diagnosis system, characterized in that: include: A collection unit is used to collect the active power and reactive power data of the total meter and sub-meters that are operating normally over a period of time; The prediction unit is used to predict the collected active power and reactive power data of the sub-meter electricity consumption by using the pre-built total meter power prediction model to obtain the predicted total meter power data, and compare the predicted total meter power data with the measured total meter power data, and make an abnormal judgment in combination with the preset monitoring rules; The diagnostic unit is used to correct the sub-meter power data using different correction methods when an abnormality is detected, and to determine the coefficient of the correction method based on minimizing the prediction error of the corrected data, and to compare the prediction residuals under different correction methods, select the correction methods with residuals from small to large as the diagnostic results, output the diagnostic results, and perform alarm processing according to the set threshold conditions.

8. The power user side line loss monitoring and diagnosis system according to claim 7 is characterized in that: The diagnostic unit is specifically used for: Establish a correction equation to correct the power data of the terminal meter, and re-input the corrected power data and the non-corrected data into the total meter power prediction model for prediction; The least square method is used to determine the best fitting coefficient of the correction equation to minimize the prediction residual after correction; Compare the prediction residuals after correction at different ends, determine the order of the line segments with abnormalities, and conduct inspections in the order of the line segments with abnormalities.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the power user side line loss monitoring and diagnosis method as described in any one of claims 1-6 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for monitoring and diagnosing line losses at the power user side as described in any one of claims 1 to 6 are implemented.

Images