A method and system for analyzing reliability of an electric energy meter

Abstract

The application provides a kind of electric energy meter reliability analysis method and system, it is related to smart meter technical field, the application is obtained in the preset time period by the reporting power curve data of each electric energy meter of total meter and each electric energy meter;Total meter power is corrected based on the preset line loss rate, and power allocation is carried out in combination with the power proportion of each electric energy meter, to construct the standard power curve data of each electric energy meter;The standard power curve data is constructed by introducing line loss correction and power allocation mechanism, and the power curve is analyzed in multiple dimensions by combining time domain and frequency domain characteristics, while the different error scales are unified by using normalization and weighted fusion method, and finally the reliability quantitative evaluation is realized by nonlinear mapping, which can effectively improve the accuracy and stability of electric energy meter measurement anomaly identification, and enhance the comparability of evaluation results under different load conditions.

Description

Technical Field

[0001] This invention relates to the field of smart meter technology, specifically to a method and system for reliability analysis of electricity meters. Background Technology

[0002] As a key terminal for power grid metering, the reliability of electricity meters directly affects the fairness of trade settlement and the accuracy of power grid operation status monitoring. With the widespread application of smart grid electricity consumption information collection systems, a large number of user electricity meters and a main meter are deployed in low-voltage distribution substations, forming a multi-level metering system. Electricity meters operate under complex electromagnetic environments and dynamic load conditions for extended periods, making them susceptible to factors such as component aging, loose wiring, and electricity theft interference, leading to metering deviations.

[0003] Current assessments of electricity meter reliability primarily rely on periodic on-site calibration, laboratory verification, or simple comparative analysis based on calculated electricity consumption over a specific period. Existing methods generally suffer from the following drawbacks: they fail to consider the impact of reasonable line losses within the distribution area on the total meter power; directly using the original total meter data as a benchmark easily introduces systematic biases; they lack modeling of the power distribution relationship of individual electricity meters under time-varying loads, making it impossible to generate a standard reference curve that matches their operating conditions; and they only focus on cumulative energy errors, ignoring the morphological changes of the power curve over time, making it difficult to distinguish between normal load fluctuations and actual metering faults.

[0004] Therefore, there is an urgent need for a method that can utilize the continuous operational data reported by the electricity meter itself to achieve multi-dimensional, real-time health status assessment and early fault warning.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for reliability analysis of electricity meters to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A reliability analysis method for electricity meters, comprising the following steps: S1: Set a preset time period, and within the preset time period, obtain the reported power curve data of the total meter in the target area and the reported power curve data of all energy meters in the area. S2: Based on the preset line loss rate of the distribution area, the power curve data reported by the total meter of the distribution area is corrected, and combined with the power ratio of the power curve data reported by each electricity meter at the corresponding time, the power allocation calibration of the power curve data reported by each electricity meter is performed to obtain the standard power curve data corresponding to each electricity meter. S3: Divide the preset time period into multiple time windows, and for each energy meter, extract the reported power curve data and standard power curve data of the energy meter within each time window; S4: For the reported power curve data and standard power curve data in each time window, calculate the difference curve between the reported power curve data and the standard power curve data, and perform integration processing on the difference curve to obtain the total energy error in that time window. At the same time, extract the derivative features and frequency domain amplitude features of the reported power curve data and standard power curve data in each time window, and determine the curve shape similarity in that time window based on the derivative features and frequency domain amplitude features. S5: Normalize and weight the total energy error and curve shape similarity within each time window to generate the metering error value of the electricity meter within the corresponding time window. S6: Iterate through the metering error values ​​of all time windows corresponding to each electricity meter, calculate the average error mapping value through the Sigmoid mapping function with a preset error threshold, and generate the reliability score of the electricity meter. Based on the reliability score, determine the metering reliability of the electricity meter.

[0008] Furthermore, the step of correcting the reported power curve data of the transformer area master table based on the preset line loss rate of the transformer area includes: By combining the reported power curve data from the transformer area master meter with a preset line loss rate, the bus line loss power curve within the transformer area is obtained. The bus line loss curve is then subtracted from the reported power curve data in the transformer area master meter to obtain the corrected power curve for the transformer area master meter. The logic for performing power sharing calibration includes: At any given moment, based on the reported power values ​​of each energy meter, the real-time power percentage of each energy meter is determined. The real-time power percentage represents the proportion of the reported power value of the corresponding energy meter in the total reported power of all energy meters in the distribution area. Based on the average reported power of each electricity meter within a preset historical time period, the historical load weight of each electricity meter is determined. The historical load weight represents the proportion of the long-term load level of the corresponding electricity meter in the long-term load level of all electricity meters in the distribution area. The real-time power ratio and the historical load weight are fused according to the preset fusion weight to obtain the comprehensive allocation coefficient of each electricity meter at the corresponding time. The fusion weight is used to characterize the degree of influence of the real-time power ratio and the historical load weight on the comprehensive allocation coefficient. Based on the comprehensive allocation coefficient of each electricity meter at the corresponding time, the corrected total power value of the distribution area is allocated to obtain the standard power value of each electricity meter at the corresponding time.

[0009] Furthermore, when the preset time period is divided into multiple time windows, each time window is continuous and non-overlapping, covering the preset time period. The logic for dividing the preset time period into multiple time windows includes: determining the power fluctuation level corresponding to each time period based on the power change rate of the power curve data between adjacent sampling times; when the power fluctuation level is greater than a preset fluctuation threshold, the corresponding time period is divided into a first time window; when the power fluctuation level is less than or equal to the preset fluctuation threshold, the corresponding time period is divided into a second time window; the length of the second time window is greater than the length of the first time window, and the multiple time windows continuously cover the preset time period.

[0010] Furthermore, the logic for extracting the derivative features of the reported power curve data and the standard power curve data is as follows: within each time window, sampling points of multiple discrete power data corresponding to the reported power curve data and the standard power curve data are obtained in chronological order to form a corresponding time series; second-order differentiation is performed on the reported power curve data and the standard power curve data to obtain the second-order derivative sequence corresponding to the reported power curve data and the second-order derivative sequence corresponding to the standard power curve data, respectively; the second-order derivative sequence is used to characterize the changing trend of the corresponding power curve at each time point; The difference between the second derivative sequence corresponding to the reported power curve data and the second derivative sequence corresponding to the standard power curve data is calculated to generate a degree of trend matching, including: Calculate the absolute difference between the second derivative of the reported power curve and the second derivative of the standard power curve at each valid sampling location within the current time window, and sum up the absolute differences to obtain the total difference between the second derivative of the reported power curve and the standard power curve. The absolute values ​​of each effective second derivative corresponding to the standard power curve data are accumulated to obtain the standard trend change intensity; Based on the relative relationship between the total difference of the second derivative values ​​and the intensity of the standard trend change, the degree of matching of the trend is determined, and the total difference of the second derivative values ​​is negatively correlated with the degree of matching of the trend.

[0011] Furthermore, the logic for obtaining the degree of matching between the internal components of the reported power curve data and the standard power curve data within each time window is as follows: Fourier transforms are performed on the reported power curve data and the standard power curve data respectively to extract their internal frequency domain components. For the frequency domain components corresponding to the standard power curve data, the frequency domain components whose amplitude values ​​exceed the first threshold are extracted, and their corresponding frequency domains are extracted as reference frequencies. Extract the amplitude of the frequency domain component corresponding to the reference frequency in the reported power curve data, and compare it with the amplitude of the frequency domain component corresponding to the reference frequency in the standard power curve data to generate the internal component matching degree, including: Calculate the absolute difference between the reported frequency domain amplitude and the standard frequency domain amplitude at each reference frequency, and sum the absolute differences to obtain the total amplitude difference. The intensity of the standard internal component is obtained by summing the frequency domain amplitudes of the standard power curve data at each reference frequency. The degree of matching of the internal components is determined based on the relative relationship between the total amplitude difference and the intensity of the standard internal components. The total amplitude difference and the degree of matching of the internal components are negatively correlated.

[0012] Furthermore, when determining the curve shape similarity based on the degree of matching of the trend of change and the degree of matching of the internal components, the degree of matching of the trend of change and the degree of matching of the internal components are weighted and fused according to a preset weight to obtain the curve shape similarity, wherein the weight coefficient corresponding to the degree of matching of the trend of change is greater than the weight coefficient corresponding to the degree of matching of the internal components.

[0013] Furthermore, the metering error value of the electricity meter within the corresponding time window is generated, including: The total energy error of the electricity meter within the current time window is normalized to obtain the normalized total energy error; The degree of deviation in curve shape is determined based on the curve shape similarity. The normalized total energy error and the deviation of the curve shape are weighted and fused according to a preset weight relationship to obtain the metering error value of the electricity meter in the current time window. The weight corresponding to the normalized total energy error is greater than the weight corresponding to the deviation of the curve shape.

[0014] Furthermore, the reliability score of the electricity meter is determined based on the metering error value within each time window, including: The average metering error value of the energy meter is obtained by averaging the metering error values ​​over all time windows. Based on the difference between the average metering error value and the preset error threshold, a reliability score for the electricity meter is generated through a preset Sigmoid mapping function. The reliability score is negatively correlated with the average metering error value.

[0015] Furthermore, the present invention also provides an electricity meter reliability analysis system for performing the above-described electricity meter reliability analysis method, comprising: Data acquisition module: Set a preset time period, and acquire the reported power curve data of the total meter of the target area and the reported power curve data of all energy meters in the area within the preset time period. Power calibration module: Based on the preset line loss rate of the distribution area, the power curve data reported by the total meter of the distribution area is corrected, and combined with the power ratio of the power curve data reported by each electricity meter at the corresponding time, the power distribution calibration of the power curve data reported by each electricity meter is performed to obtain the standard power curve data corresponding to each electricity meter. Time window segmentation module: Divides the preset time period into multiple time windows, and extracts the reported power curve data and standard power curve data of the energy meter within each time window for each energy meter; Feature Analysis Module: For the reported power curve data and standard power curve data within each time window, calculate the difference curve between the reported power curve data and the standard power curve data, and perform integration processing on the difference curve to obtain the total energy error within that time window. At the same time, extract the derivative features and frequency domain amplitude features of the reported power curve data and standard power curve data within each time window, and determine the curve shape similarity within that time window based on the derivative features and frequency domain amplitude features. Metering error calculation module: Normalizes and weights the total energy error and curve shape similarity within each time window to generate the metering error value of the electricity meter within the corresponding time window; Reliability assessment module: It iterates through the metering error values ​​of all time windows corresponding to each electricity meter, calculates the average error mapping value through the Sigmoid mapping function with a preset error threshold, and generates a reliability score for the electricity meter. The metering reliability is judged based on the reliability score.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention acquires the power curves of the main meter and all user electricity meters in a distribution area, corrects the main meter data based on a preset line loss rate, and then performs apportionment calibration by combining the power proportion of each electricity meter at the corresponding time. This generates a standard power curve for each electricity meter that matches its actual load characteristics, solving the benchmark distortion problem caused by neglecting the distribution area topology and line loss in existing technologies. By dividing the time period into multiple windows, the reported power curve and the standard power curve are compared and analyzed from three dimensions: energy error, derivative characteristics, and frequency domain amplitude characteristics within each time window. This not only characterizes the overall metering deviation but also reflects the curve's changing trend and internal structural differences, thereby improving the accuracy of anomaly identification. Furthermore, by normalizing the total energy error and weighting and fusing it with curve shape similarity, a unified-scale metering error index is constructed. This is then combined with the Sigmoid mapping function to generate a continuous reliability score, achieving an effective mapping from multi-dimensional features to a single evaluation index, enabling good comparability and hierarchical judgment capabilities between different electricity meters. It effectively quantifies the degree of deviation of the curve in dynamic shape, overcoming the limitation of traditional methods that rely solely on accumulated power and cannot identify transient or waveform-level anomalies. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a schematic diagram illustrating the extraction of power curve data reported in this invention; Figure 3 This is a schematic diagram of the time window division in this invention; Figure 4 This is a schematic diagram of the multi-feature analysis process within each time window of the present invention; Figure 5 This is a system structure diagram of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0019] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0020] Example: To facilitate understanding of this application, the embodiments of this application will be briefly described below: This embodiment provides a method for reliability analysis of electricity meters, applicable to the assessment of the operating status of distribution area-level electricity meters in an electricity consumption information collection system. Based on the power curve data reported by the distribution area's main meter and individual electricity meters, this method constructs standard power curve data and performs comparative analysis based on multi-dimensional features to achieve a quantitative assessment of the metering reliability of the electricity meters.

[0021] Please see Figure 1 The present invention provides a technical solution: A reliability analysis method for electricity meters, comprising the following steps: S1: Set a preset time period, and within the preset time period, obtain the reported power curve data of the total meter in the target area and the reported power curve data of all energy meters in the area. Please see Figure 2 This is a schematic diagram of power curve data extraction provided in an embodiment of this application. In this embodiment, the preset time period is used to perform statistical analysis on the electricity meter's operating data, and its length is set according to actual application requirements. Since the electricity consumption data of the electricity meter usually has periodic variation characteristics, it is preferable to set the preset time period to a time range that can cover one or more complete electricity consumption cycles, such as 1 day, 3 days, or 7 days.

[0022] When the preset time period is less than a complete electricity consumption cycle, it may not fully reflect the user's electricity consumption patterns and is easily affected by short-term fluctuations. When the preset time period is too long, it increases the complexity of data processing and reduces the sensitivity to recent abnormal behavior of the electricity meter. Therefore, in this embodiment, by reasonably setting the preset time period, the data is both representative and ensures the real-time nature and stability of the analysis results.

[0023] The reported power curve data is time-series data collected according to a preset sampling period, with each moment corresponding to a power sample value. The power curve can be visualized by connecting adjacent sampling points, but it is processed as discrete data points during the calculation process. In some embodiments, the discrete sampling points can be interpolated to obtain a smoother curve representation, but this invention does not rely on the interpolation process.

[0024] S2: Based on the preset line loss rate of the distribution area, the power curve data reported by the total meter of the distribution area is corrected, and combined with the power ratio of the power curve data reported by each electricity meter at the corresponding time, the power allocation calibration of the power curve data reported by each electricity meter is performed to obtain the standard power curve data corresponding to each electricity meter. In a transformer substation power supply system, the electrical energy measured by the substation master meter includes not only the electricity consumption of each user's meter but also the losses generated during transmission by the distribution lines and related equipment, i.e., line losses. Therefore, the power value of the substation master meter is usually greater than the sum of the power values ​​of each individual meter. These line losses mainly originate from conductor resistance loss, transformer loss, and contact loss, and their magnitude is related to factors such as load current, line length, and operating conditions. In practical engineering, a preset line loss rate is usually set for each substation based on historical operating data or empirical models to describe the average loss level of the substation under normal operating conditions. Without considering line losses, directly using the substation master meter power as a reference for allocation or comparison will lead to an overall higher reference value for all meters, introducing a systematic bias and affecting the accuracy of subsequent metering error analysis. Therefore, this embodiment first corrects the substation master meter power curve based on the preset line loss rate. The power curve data reported by the transformer substation's master meter is corrected based on the preset line loss rate for that substation, including: By combining the power curve data reported by the transformer area master meter with the preset line loss rate, the bus loss power curve within the transformer area is obtained. The bus loss curve is then subtracted from the power curve data reported by the transformer area master meter to obtain the corrected power curve of the transformer area master meter. After obtaining the corrected total power curve for the distribution area, in order to further construct a reference benchmark for each electricity meter, it is necessary to reasonably allocate the total power. The logic for power allocation calibration includes: at each moment, based on the reported power value of each energy meter, determining the real-time power ratio of each energy meter, wherein the real-time power ratio represents the proportion of the reported power value of the corresponding energy meter in the total reported power of all energy meters in the distribution area; Based on the average reported power of each electricity meter within a preset historical time period, the historical load weight of each electricity meter is determined. The historical load weight represents the proportion of the long-term load level of the corresponding electricity meter in the long-term load level of all electricity meters in the distribution area. The real-time power ratio and the historical load weight are fused according to the preset fusion weight to obtain the comprehensive allocation coefficient of each electricity meter at the corresponding time. The fusion weight is used to characterize the degree of influence of the real-time power ratio and the historical load weight on the comprehensive allocation coefficient. Based on the comprehensive allocation coefficient of each electricity meter at the corresponding time, the corrected total power value of the distribution area is allocated to obtain the standard power value of each electricity meter at the corresponding time. Based on the standard power values ​​at various times, standard power curve data corresponding to each electricity meter are constructed; the formula is as follows: in, Indicates the first Each electricity meter at time The overall allocation coefficient for the total power of the modified background area is adjusted. This indicates that the q-th energy meter is at time q. Standard power value, Indicates at time Correct the total power value of the background area. Indicates the first Each electricity meter at time The reported power value, For the first Historical load weight of each electricity meter No. The average reported power of each electricity meter within a preset historical time period, where M represents the total number of electricity meters in the distribution area. This is a preset value, representing the relationship between the real-time power ratio and the historical load weight in the calculation of the comprehensive allocation coefficient. The corresponding weight of the real-time power percentage item, The weights corresponding to the historical load weighting items; When μ is larger, the allocation result relies more on the power data at the current moment, which can improve the responsiveness to real-time load changes; when μ is smaller, the allocation result relies more on historical load characteristics, which helps to improve the stability of the allocation result. Experimental verification shows that when μ is between 0.6 and 0.8, a better balance can be achieved between real-time performance and stability. In this embodiment, μ is set to 0.7.

[0025] Compared with methods that use fixed allocation coefficients or historical average ratios, this allocation method based on real-time power proportions can reflect the actual load changes of different energy meters in different time periods, making the constructed standard power curve data have higher dynamic consistency and accuracy. Through the above-described line loss correction and power allocation calibration process, this embodiment can construct a standard power curve data for each electricity meter that matches its actual operating state, thereby providing a reliable reference benchmark for subsequent metering error analysis based on curve differences and avoiding misjudgment problems caused by line loss and uneven load distribution.

[0026] S3: Divide the preset time period into multiple time windows, and for each energy meter, extract the reported power curve data and standard power curve data of the energy meter within each time window; Please see Figure 3 This is a schematic diagram of time window division provided in an embodiment of this application. In this embodiment, the time window is divided into preset time periods according to a fixed length. Each time window is contiguous and non-overlapping, thus continuously covering the entire preset time period. Within the preset time period, the time windows can be divided into equal-length segments. For example, when the preset time period is 1 day (24 hours), the preferred time window length is 1 hour, dividing the preset time period into 24 time windows. These time windows are sequentially connected and cover the entire time period, thus avoiding data loss, duplicate calculation of the same data, or non-reproducible time window division. In other embodiments, the time windows can be dynamically divided. Based on the power change rate of the power curve data between adjacent sampling times, the power fluctuation level corresponding to each time period is determined. When the power fluctuation level is greater than a preset fluctuation threshold, the corresponding time period is divided into a shorter time window; when the power fluctuation level is less than or equal to the preset fluctuation threshold, the corresponding time period is divided into a longer time window. The logic is that using a longer time window during periods of stable load helps improve statistical stability and reduce interference from short-term fluctuations; using a shorter time window during periods of significant load fluctuation helps improve the response speed to abnormal changes, thus balancing evaluation accuracy and computational efficiency.

[0027] S4: For the reported power curve data and standard power curve data in each time window, calculate the difference curve between the reported power curve data and the standard power curve data, and integrate the difference curve to obtain the total energy error in that time window. At the same time, extract the derivative features and frequency domain amplitude features of the reported power curve data and the standard power curve data respectively, and determine the curve shape similarity based on the derivative features and frequency domain amplitude features. Please see Figure 4 This is a schematic diagram of the multi-feature analysis process within a time window provided in the embodiments of this application. The logic for extracting the derivative features of the reported power curve data and the standard power curve data is as follows: Within each time window, sampling points of multiple discrete power data corresponding to the reported power curve data and the standard power curve data are extracted in chronological order to form a corresponding time series. In this embodiment, the time series data originates from the reported power curve data collected in step S1 according to a preset sampling period; the discrete power data sampling points within the time window are a subset of sampling points selected from the reported power curve data that fall within the corresponding time window range; the sampling period is less than the time window length, so that each time window contains multiple sampling points, preferably, each time window contains no less than 4 sampling points to meet the needs of subsequent calculations; The second derivatives of the reported power curve data and the standard power curve data are performed to obtain the second derivative sequences corresponding to the reported power curve data and the standard power curve data, respectively; the second derivative sequences are used to characterize the changing trend of the corresponding power curves at each time step. The second derivative is calculated using a discrete difference method; for a discrete power sequence with a fixed time interval... The second derivative at its i-th sampling point can be calculated using the following formula: in, This represents the power value at the i-th sampling point. This represents the power value at the (i+1)th sampling point. This represents the power value at the (i-1)th sampling point; In this embodiment, the second derivative is calculated independently based on the discrete power data points within each time window, and no cross-window data references are made between time windows; for each sampling point within a time window, the second derivative is calculated only for the intermediate sampling point with adjacent sampling points, and the sampling points at the window boundary do not participate in the second derivative calculation. The difference between the second derivative sequence corresponding to the reported power curve data and the second derivative sequence corresponding to the standard power curve data is calculated to generate a degree of matching in terms of trend; including: Calculate the absolute difference between the second derivative of the reported power curve and the second derivative of the standard power curve at each valid sampling location within the current time window, and sum up the absolute differences to obtain the total difference between the second derivative of the reported power curve and the standard power curve. The absolute values ​​of each effective second derivative corresponding to the standard power curve data are accumulated to obtain the standard trend change intensity; Based on the relative relationship between the total difference of the second derivative values ​​and the intensity of the standard trend change, the degree of matching of the trend is determined, and the total difference of the second derivative values ​​is negatively correlated with the degree of matching of the trend; the formula used is: in, Indicates the degree of matching of trends of change. This represents the i-th effective second derivative value corresponding to the reported power curve data. This represents the i-th effective second derivative value corresponding to the standard power curve data, and m represents the number of effective second derivative values ​​within the current time window. This is a preset non-zero minimum value to avoid calculation instability caused by a denominator that is zero or too small. The preferred value range is [value range missing]. .

[0028] The logic for obtaining the degree of matching between the reported power curve data and the standard power curve data within each time window is as follows: Fourier transforms are performed on the reported power curve data and the standard power curve data, respectively. In this embodiment, specifically, discrete Fourier transforms are performed on the discrete power data sequences corresponding to the reported power curve data and the standard power curve data to extract their internal frequency domain components. For the frequency domain components corresponding to the standard power curve data, frequency domain components with amplitude values ​​exceeding a first threshold are extracted, and their corresponding frequency domains are extracted as reference frequencies. In this embodiment, the first threshold is used to screen representative frequency domain components in the standard power curve data, and its value can be adaptively set according to the frequency domain amplitude distribution. Preferably, the first threshold is 20% to 50% of the maximum amplitude. Extract the amplitude of the frequency domain component corresponding to the reference frequency in the reported power curve data, and compare it with the amplitude of the frequency domain component corresponding to the reference frequency in the standard power curve data to generate the degree of matching of internal components, including: Calculate the absolute difference between the reported frequency domain amplitude and the standard frequency domain amplitude at each reference frequency, and sum the absolute differences to obtain the total amplitude difference. The intensity of the standard internal component is obtained by summing the frequency domain amplitudes of the standard power curve data at each reference frequency. Based on the relative relationship between the total amplitude difference and the intensity of the standard internal components, the degree of matching of the internal components is determined, and the total amplitude difference and the degree of matching of the internal components are negatively correlated; the formula used is: in, Indicates the degree of matching of internal components. This represents the frequency domain amplitude of the reported power curve data at the k-th reference frequency. This represents the frequency domain amplitude of the standard power curve data at the k-th reference frequency, where k represents the number of reference frequencies. It is a preset non-zero minimum value.

[0029] The This is a preset non-zero minimum value used to avoid calculation instability caused by a denominator that is zero or too small. Preferably, its value range is... Furthermore, it can be adaptively set according to the magnitude of the frequency domain amplitude.

[0030] The curve shape similarity is determined based on the matching degree of the trend of change and the matching degree of the internal components. The matching degree of the trend of change and the matching degree of the internal components are then weighted and fused according to preset weights to obtain the curve shape similarity. The weight coefficient corresponding to the matching degree of the trend of change is greater than the weight coefficient corresponding to the matching degree of the internal components. The formula used is as follows: in, Indicates the similarity of curve shapes. Indicates the degree of matching between the curve trends. Indicates the degree of matching of internal components. , These represent the degree of matching in terms of curve trend and the degree of matching in terms of internal components, respectively, with preset weighting coefficients. and .

[0031] Preferably, the weighting coefficient is taken as follows: =0.6, =0.4. Specifically, by selecting historical power data from normal and abnormal energy meters in multiple distribution areas, the similarity of curve shapes was calculated under different weight combinations, and the differentiation effect of subsequent metering error values ​​and reliability scores on abnormal energy meters was compared. Experimental results show that when taking... =0.6, When the weighting factor is 0.4, it can adequately reflect the sensitivity of the trend matching degree to metering anomalies, while also taking into account the characterization effect of the internal frequency domain component matching degree on the differences in periodic fluctuations. This results in a higher differentiation between normal and abnormal energy meters in reliability scoring. This is because the trend matching degree, based on the second derivative characteristic, can more directly reflect the changing trend and abrupt changes in the power curve, while the internal component matching degree, based on frequency domain characteristics, mainly reflects the periodic structure. Therefore, the trend matching degree is given a higher weight. In other embodiments, the weighting coefficient can be adjusted according to actual needs.

[0032] In this embodiment, the total energy error is used to characterize the cumulative deviation between the reported power curve data and the standard power curve data in the time dimension. It is obtained by discretely integrating the power difference curve. Specifically, at each sampling moment, the power difference between the two is calculated, and the absolute value of the difference is multiplied by the sampling time interval to obtain the energy error within that time period. The energy errors at each sampling moment within the time window are accumulated to obtain the total energy error. The formula for calculating the total energy error is as follows: in, For total energy error, Let be the power difference at the i-th sampling time. Report the power curve data at the power value at the i-th sampling time. This represents the power value of the standard power curve data at the i-th sampling time, where n is the total number of sampling times within the window. The sampling period.

[0033] S5: Normalize and weight the total energy error and curve shape similarity within each time window to generate the metering error value of the electricity meter within the corresponding time window. Generate the metering error value of the electricity meter within the corresponding time window, including: The total energy error of the electricity meter within the current time window is normalized to obtain the normalized total energy error; The degree of deviation in curve shape is determined based on the curve shape similarity. The normalized total energy error and the deviation of the curve shape are weighted and fused according to a preset weighting relationship to obtain the metering error value of the energy meter within the current time window. The weight corresponding to the normalized total energy error is greater than the weight corresponding to the deviation of the curve shape; the formula is as follows: in, This indicates the metering error value of the electricity meter within the current time window; The result after normalizing the total energy error. Indicates the similarity of curve shapes. , Indicates the corresponding weight. and + =1. Experimental analysis determined that the optimal weighting coefficient is [value missing]. =0.6、 =0.4.

[0034] because The value ranges from 0 to 1. To ensure dimensional consistency, the total energy error needs to be normalized. The formula for normalizing the total energy error is as follows: in, The total energy error within the corresponding time window, The total energy error after normalization. The energy value is the standard power curve data within the corresponding time window. This is a preset non-zero minimum value to avoid calculation instability caused by a denominator that is zero or too small. The preferred value range is [value range missing]. .

[0035] S6: Iterate through the metering error values ​​of all time windows corresponding to each electricity meter, calculate the average error mapping value through the Sigmoid mapping function with a preset error threshold, and generate the reliability score of the electricity meter. Based on the reliability score, determine the metering reliability of the electricity meter.

[0036] After obtaining the metering error values ​​within each time window, in order to comprehensively evaluate the metering reliability of the electricity meter over the entire preset time period, it is necessary to summarize and process the errors from multiple time windows.

[0037] Specifically, for each electricity meter, all its corresponding time windows are traversed, and the metering error values ​​within each time window are averaged to obtain the average metering error, based on the following formula: in: This represents the metering error value in the j-th time window, where N is the number of time windows. Using an averaging method comprehensively reflects the overall error level of the electricity meter throughout the entire time period, while avoiding excessive influence of abnormal fluctuations in a single time window on the final evaluation result, thus improving the stability of the evaluation results.

[0038] Generating the reliability score includes: averaging the metering error values ​​of the energy meter across all time windows to obtain an average metering error value; and generating a reliability score for the energy meter based on the difference between the average metering error value and a preset error threshold using a preset Sigmoid mapping function. The reliability score is negatively correlated with the average metering error value, according to the following formula: in, This represents the average value of the metering error of the electricity meter over all time windows. This represents the preset error threshold, and 'a' represents the slope adjustment coefficient of the Sigmoid mapping function, where a > 0. This indicates the reliability score of the electricity meter; The lower the reliability score, the lower the corresponding measurement reliability.

[0039] In this embodiment, preferably, the preset error threshold is taken as... =0.3, the slope adjustment coefficient of the Sigmoid mapping function is taken as... = 10, of which, =0.3 is used to distinguish between normal and abnormal metering states of the electricity meter. When the average metering error is below this threshold, the reliability score is higher; when the average metering error is above this threshold, the reliability score is lower. Parameter = 10 is used to control the sensitivity of the reliability score to changes around a threshold, so that the score results can effectively distinguish critical error states while maintaining a certain degree of stability. In other embodiments, the preset error threshold... The slope adjustment coefficient can be selected within the range of 0.2 to 0.4. You can select from the range of 5 to 15.

[0040] Based on the above reliability scoring formula, when = When R=0.5, preferably, when R>0.5, the meter is considered to have high metering reliability, and when R<0.5, the meter is considered to have a risk of decreased metering reliability.

[0041] Please see Figure 5 The present invention also provides an energy meter reliability analysis system for performing the above-described energy meter reliability analysis method, comprising: Data acquisition module: Set a preset time period, and acquire the reported power curve data of the total meter of the target area and the reported power curve data of all energy meters in the area within the preset time period. Power calibration module: Based on the preset line loss rate of the distribution area, the power curve data reported by the total meter of the distribution area is corrected, and combined with the power ratio of the power curve data reported by each electricity meter at the corresponding time, the power distribution calibration of the power curve data reported by each electricity meter is performed to obtain the standard power curve data corresponding to each electricity meter. Time window segmentation module: Divides the preset time period into multiple time windows, and extracts the reported power curve data and standard power curve data of the energy meter within each time window for each energy meter; Feature Analysis Module: For the reported power curve data and standard power curve data within each time window, calculate the difference curve between the reported power curve data and the standard power curve data, and integrate the difference curve to obtain the total energy error within that time window. At the same time, extract the derivative features and frequency domain amplitude features of the reported power curve data and the standard power curve data respectively, and determine the curve shape similarity based on the derivative features and frequency domain amplitude features. Metering error calculation module: Normalizes and weights the total energy error and curve shape similarity within each time window to generate the metering error value of the electricity meter within the corresponding time window; Reliability assessment module: It iterates through the metering error values ​​of all time windows corresponding to each electricity meter, calculates the average error mapping value through the Sigmoid mapping function with a preset error threshold, and generates a reliability score for the electricity meter. The metering reliability is judged based on the reliability score.

[0042] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0043] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0044] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0045] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for reliability analysis of electricity meters, characterized in that, The specific steps include: S1: Set a preset time period, and within the preset time period, obtain the reported power curve data of the total meter in the target area and the reported power curve data of all energy meters in the area. S2: Based on the preset line loss rate of the distribution area, the power curve data reported by the total meter of the distribution area is corrected, and combined with the power ratio of the power curve data reported by each electricity meter at the corresponding time, the power allocation calibration of the power curve data reported by each electricity meter is performed to obtain the standard power curve data corresponding to each electricity meter. S3: Divide the preset time period into multiple time windows, and for each energy meter, extract the reported power curve data and standard power curve data of the energy meter within each time window; S4: For the reported power curve data and standard power curve data in each time window, calculate the difference curve between the reported power curve data and the standard power curve data, and perform integration processing on the difference curve to obtain the total energy error in that time window. At the same time, extract the derivative features and frequency domain amplitude features of the reported power curve data and standard power curve data in each time window, and determine the curve shape similarity in that time window based on the derivative features and frequency domain amplitude features. S5: Normalize and weight the total energy error and curve shape similarity within each time window to generate the metering error value of the electricity meter within the corresponding time window. S6: Iterate through the metering error values ​​of all time windows corresponding to each electricity meter, calculate the average error mapping value through the Sigmoid mapping function with a preset error threshold, and generate the reliability score of the electricity meter. Based on the reliability score, determine the metering reliability of the electricity meter.

2. The method for reliability analysis of an electricity meter according to claim 1, characterized in that, The correction of the reported power curve data of the transformer area master table based on the preset line loss rate of the transformer area includes: By combining the reported power curve data from the transformer area master meter with a preset line loss rate, the bus line loss power curve within the transformer area is obtained. The bus line loss curve is then subtracted from the reported power curve data in the transformer area master meter to obtain the corrected power curve for the transformer area master meter. The logic for performing power sharing calibration includes: At any given moment, based on the reported power values ​​of each energy meter, the real-time power percentage of each energy meter is determined. The real-time power percentage represents the proportion of the reported power value of the corresponding energy meter in the total reported power of all energy meters in the distribution area. Based on the average reported power of each electricity meter within a preset historical time period, the historical load weight of each electricity meter is determined. The historical load weight represents the proportion of the long-term load level of the corresponding electricity meter in the long-term load level of all electricity meters in the distribution area. The real-time power ratio and the historical load weight are fused according to the preset fusion weight to obtain the comprehensive allocation coefficient of each electricity meter at the corresponding time. The fusion weight is used to characterize the degree of influence of the real-time power ratio and the historical load weight on the comprehensive allocation coefficient. Based on the comprehensive allocation coefficient of each electricity meter at the corresponding time, the corrected total power value of the distribution area is allocated to obtain the standard power value of each electricity meter at the corresponding time. Based on the standard power values ​​at each time point, standard power curve data corresponding to each electricity meter is constructed.

3. The method for reliability analysis of an electricity meter according to claim 1, characterized in that, When a preset time period is divided into multiple time windows, each time window is continuous and non-overlapping, covering the preset time period. The logic for dividing the preset time period into multiple time windows includes: determining the power fluctuation level corresponding to each time period based on the power change rate of power curve data between adjacent sampling times; when the power fluctuation level is greater than a preset fluctuation threshold, the corresponding time period is divided into a first time window; when the power fluctuation level is less than or equal to the preset fluctuation threshold, the corresponding time period is divided into a second time window; the length of the second time window is greater than the length of the first time window, and the multiple time windows continuously cover the preset time period.

4. The method for reliability analysis of an electricity meter according to claim 1, characterized in that, The logic for extracting the derivative features of the reported power curve data and the standard power curve data is as follows: Within each time window, sampling points of multiple discrete power data corresponding to the reported power curve data and the standard power curve data are obtained in chronological order to form a corresponding time series; second-order differentiation is performed on the reported power curve data and the standard power curve data to obtain the second-order derivative sequence corresponding to the reported power curve data and the second-order derivative sequence corresponding to the standard power curve data, respectively; the second-order derivative sequence is used to characterize the changing trend of the corresponding power curve at each time point; The difference between the second derivative sequence corresponding to the reported power curve data and the second derivative sequence corresponding to the standard power curve data is calculated to generate a degree of trend matching, including: Calculate the absolute difference between the second derivative of the reported power curve and the second derivative of the standard power curve at each valid sampling location within the current time window, and sum up the absolute differences to obtain the total difference between the second derivative of the reported power curve and the standard power curve. The absolute values ​​of each effective second derivative corresponding to the standard power curve data are accumulated to obtain the standard trend change intensity; Based on the relative relationship between the total difference of the second derivative values ​​and the intensity of the standard trend change, the degree of matching of the trend is determined, and the total difference of the second derivative values ​​is negatively correlated with the degree of matching of the trend.

5. The method for reliability analysis of an electricity meter according to claim 4, characterized in that, The logic for obtaining the degree of matching between the reported power curve data and the standard power curve data within each time window is as follows: Fourier transforms are performed on the reported power curve data and the standard power curve data respectively to extract their internal frequency domain components. For the frequency domain components corresponding to the standard power curve data, the frequency domain components whose amplitude values ​​exceed the first threshold are extracted, and their corresponding frequency domains are extracted as reference frequencies. Extract the amplitude of the frequency domain component corresponding to the reference frequency in the reported power curve data, and compare it with the amplitude of the frequency domain component corresponding to the reference frequency in the standard power curve data to generate the internal component matching degree, including: Calculate the absolute difference between the reported frequency domain amplitude and the standard frequency domain amplitude at each reference frequency, and sum the absolute differences to obtain the total amplitude difference. The intensity of the standard internal component is obtained by summing the frequency domain amplitudes of the standard power curve data at each reference frequency. The degree of matching of the internal components is determined based on the relative relationship between the total amplitude difference and the intensity of the standard internal components. The total amplitude difference and the degree of matching of the internal components are negatively correlated.

6. The method for reliability analysis of an electricity meter according to claim 4, characterized in that, When determining the curve shape similarity based on the degree of matching of the trend and the degree of matching of the internal components, the degree of matching of the trend and the degree of matching of the internal components are weighted and fused according to a preset weight to obtain the curve shape similarity, wherein the weight coefficient corresponding to the degree of matching of the trend is greater than the weight coefficient corresponding to the degree of matching of the internal components.

7. The method for reliability analysis of an electricity meter according to claim 6, characterized in that, Generate the metering error value of the electricity meter within the corresponding time window, including: The total energy error of the electricity meter within the current time window is normalized to obtain the normalized total energy error; The degree of deviation in curve shape is determined based on the curve shape similarity. The normalized total energy error and the deviation of the curve shape are weighted and fused according to a preset weight relationship to obtain the metering error value of the electricity meter in the current time window. The weight corresponding to the normalized total energy error is greater than the weight corresponding to the deviation of the curve shape.

8. The method for reliability analysis of an electricity meter according to claim 7, characterized in that, The reliability score of the electricity meter is determined based on the metering error value within each time window, including: The average metering error value of the energy meter is obtained by averaging the metering error values ​​over all time windows. Based on the difference between the average metering error value and the preset error threshold, a reliability score for the electricity meter is generated through a preset Sigmoid mapping function. The reliability score is negatively correlated with the average metering error value.

9. A reliability analysis system for electricity meters, characterized in that, The electricity meter reliability analysis system is used to execute the electricity meter reliability analysis method according to any one of claims 1-8, including: Data acquisition module: Set a preset time period, and acquire the reported power curve data of the total meter of the target area and the reported power curve data of all energy meters in the area within the preset time period. Power calibration module: Based on the preset line loss rate of the distribution area, the power curve data reported by the total meter of the distribution area is corrected, and combined with the power ratio of the power curve data reported by each electricity meter at the corresponding time, the power distribution calibration of the power curve data reported by each electricity meter is performed to obtain the standard power curve data corresponding to each electricity meter. Time window segmentation module: Divides the preset time period into multiple time windows, and extracts the reported power curve data and standard power curve data of the energy meter within each time window for each energy meter; Feature Analysis Module: For the reported power curve data and standard power curve data within each time window, calculate the difference curve between the reported power curve data and the standard power curve data, and perform integration processing on the difference curve to obtain the total energy error within that time window. At the same time, extract the derivative features and frequency domain amplitude features of the reported power curve data and standard power curve data within each time window, and determine the curve shape similarity within that time window based on the derivative features and frequency domain amplitude features. Metering error calculation module: Normalizes and weights the total energy error and curve shape similarity within each time window to generate the metering error value of the electricity meter within the corresponding time window; Reliability assessment module: It iterates through the metering error values ​​of all time windows corresponding to each electricity meter, calculates the average error mapping value through the Sigmoid mapping function with a preset error threshold, and generates a reliability score for the electricity meter. The metering reliability is judged based on the reliability score.

Images