Intelligent meter box operation anomaly detection method and system

Abstract

This application relates to the field of metering box anomaly detection technology, and discloses a method and system for detecting operational anomalies in intelligent metering boxes. The method includes: collecting measurement data from each sub-meter connected to the same energy metering box at predetermined time intervals; constructing a data sequence for each sub-meter based on the measurement data acquired over multiple consecutive time intervals; simultaneously collecting power data from each sub-meter while collecting its measurement data; identifying whether the power data of each sub-meter is less than a preset power threshold based on the power data; if it is lower than the power threshold, determining whether a measurement anomaly exists; if so, marking the measurement data at the corresponding acquisition time as invalid; calculating the coupling degree between any data sequence and other data sequences, setting a coupling threshold, and determining anomalies based on the calculated coupling degree and coupling threshold. This invention can improve the reliability and sensitivity of metering box anomaly detection.

Description

Technical Field

[0001] This application relates to the field of meter box anomaly detection technology, and in particular to a method and system for detecting abnormal operation of an intelligent meter box. Background Technology

[0002] Electricity metering boxes are typically connected to multiple electricity meters to measure the electricity consumption of different users within the same power supply circuit. Under normal operating conditions, the voltage, frequency, and other electrical quantities measured by each meter should be highly consistent. When a meter malfunctions, its wiring becomes loose, or its external current transformer malfunctions, its measurement data will deviate from the true value, leading to inaccurate metering. Therefore, detecting abnormalities in the operating status of the metering box is of great importance.

[0003] Currently, anomaly detection in meter boxes mainly relies on the following methods: First, regular manual inspections, which are inefficient, costly, and difficult to detect early or intermittent faults. Second, self-testing based on individual meters, such as detecting power status or internal register errors, but this cannot detect measurement deviations inconsistent with other meters. Third, comparing the difference or ratio of measurements from two meters to determine anomalies; however, this method is susceptible to synchronization errors at sampling time, instantaneous noise, and load fluctuations, resulting in a high false alarm rate.

[0004] In recent years, some research has applied artificial intelligence methods such as machine learning and deep learning to anomaly detection in electricity metering equipment. For example, classification or prediction models are trained by collecting large amounts of historical data to identify abnormal patterns. While these methods have shown some effectiveness on specific datasets, they have significant limitations: they require a large number of well-labeled anomaly samples, while in actual operation, anomaly data is scarce and diverse; the model training process is complex and computationally resource-intensive; and the model is prone to failure when the electrical environment or load characteristics of the metering box change, requiring retraining. Furthermore, prediction-based methods typically rely on long-term series prediction errors to identify anomalies, lacking sensitivity to sudden changes. Finally, these methods have high computational resource requirements, making them difficult to deploy on edge devices and resulting in high deployment costs.

[0005] To address the aforementioned problems, this invention provides a method and system for detecting operational anomalies in smart metering boxes. Without requiring pre-trained models, it calculates the coupling degree of measurement data from individual meters connected to the same metering box, and combines this with power data to eliminate light-load interference. Anomaly detection is then performed based on the coupling degree and coupling threshold. Compared to machine learning and deep learning methods, this solution is more suitable for deployment in field terminal devices such as metering boxes. Summary of the Invention

[0006] This application provides a method and system for detecting abnormal operation of intelligent metering boxes, which improves the reliability and sensitivity of metering box abnormality detection.

[0007] Firstly, this application provides a method for detecting abnormal operation of an intelligent metering box, the method comprising: S1. Collect measurement data from each sub-meter connected to the same energy metering box at predetermined time intervals. For each sub-meter, construct a data sequence based on the measurement data acquired at multiple consecutive time intervals. S2. While collecting the measurement data of each sub-meter, simultaneously collect the power data of each sub-meter. Based on the power data, identify whether the power data of each sub-meter is less than the preset power threshold. If it is less than the power threshold, determine whether there is a measurement abnormality. If so, mark the measurement data at the corresponding collection time as invalid. S3. For any data sequence, calculate the degree of coupling with other data sequences, set a coupling threshold, and judge anomalies based on the calculated degree of coupling and the coupling threshold.

[0008] In conjunction with the first aspect, in the first implementation of the first aspect of this application, the anomaly determination based on the calculated coupling degree and coupling threshold includes: S31. If the coupling degree between all pairs of data sequences is greater than or equal to the coupling threshold, the metering box is judged to be operating normally. Otherwise, if there is a data sequence whose coupling degree with all other data sequences is less than the coupling threshold, the device corresponding to the data sequence is judged to be operating abnormally. S32. Otherwise, it indicates that there exists a set of data sequences consisting of at least two data sequences. Within the set of data sequences, the coupling degree between any two data sequences is less than the coupling threshold, and the coupling degree between any data sequence within the set and a data sequence outside the set is partially less than or partially greater than or equal to the coupling threshold. In this case, it is determined that there is a group anomaly.

[0009] In conjunction with the first aspect, in the second implementation of the first aspect of this application, the degree of coupling between computation and other data sequences includes: The current data sequence is referred to as the first data sequence. When calculating the coupling degree between the first data sequence and any other data sequence, the first data sequence is translated on the time axis relative to the other data sequences, and the coupling degree between the translated first data sequence and other data sequences is calculated.

[0010] In conjunction with the first aspect, in the third implementation of the first aspect of this application, calculating the coupling degree between the translated first data sequence and other data sequences includes: Other data sequences are referred to as the second data sequence. When the first data sequence is translated relative to the second data sequence on the time axis, multiple translations with different amounts are performed, and the coupling degree of the first and second data sequences under multiple translations is calculated to obtain multiple coupling degrees. The maximum coupling degree is taken as the final coupling degree of the first data sequence relative to the second data sequence.

[0011] In conjunction with the first aspect, in the fourth implementation of the first aspect of this application, a coupling threshold is set, including: Get the deployment time of the electricity metering box and each sub-meter. If the deployment time is less than or equal to the preset first time, set the coupling threshold to the first threshold. If the deployment duration is greater than or equal to the preset first duration, the coupling threshold is calculated based on historical data every preset time period.

[0012] In conjunction with the first aspect, in the fifth implementation of the first aspect of this application, the coupling threshold is calculated based on historical data, including: Obtain the time period during which all devices are running normally in the recent period, calculate the degree of coupling between the data sequences of all sub-tables within the time period, calculate the mean and standard deviation of all coupling degrees, and set the coupling threshold as the mean minus a preset multiple multiplied by the standard deviation.

[0013] In conjunction with the first aspect, in the sixth implementation of the first aspect of this application, after calculating the coupling threshold, the following is also performed: Record the trend of the coupling degree between each data sequence and all other data sequences over time. When the coupling degree of a certain data sequence shows a monotonically decreasing trend over multiple consecutive sampling periods, and the current coupling degree is still higher than the coupling threshold, it is determined that the device corresponding to the data sequence has early abnormal signs.

[0014] In conjunction with the first aspect, in the seventh implementation of the first aspect of this application, determining whether a measurement anomaly exists includes: For each sub-table, obtain the measurement data corresponding to all the collection times marked as invalid, calculate the fluctuation of the measurement data at the invalid times, and at the same time calculate the fluctuation of other sub-tables at the same collection times. When the difference between the fluctuation of a certain sub-table and the average fluctuation of all other sub-tables is greater than the preset fluctuation threshold, it is determined that there is a measurement anomaly in the corresponding sub-table.

[0015] Secondly, this application provides an intelligent metering box malfunction detection system, the system comprising: The data acquisition module is used to collect measurement data from each sub-meter in the same energy metering box at predetermined time intervals. For each sub-meter, a data sequence is constructed based on the measurement data acquired at multiple consecutive time intervals. The data filtering module is used to collect power data of each sub-meter while collecting measurement data of each sub-meter. Based on the power data, it identifies whether the power data of each sub-meter is less than the preset power threshold. If it is less than the power threshold, the measurement data at the corresponding collection time is marked as invalid, and all data points collected at the collection time of invalid measurement data in all data sequences are removed. The anomaly detection module is used to calculate the degree of coupling between any given data sequence and other data sequences, set a coupling threshold, and determine anomalies based on the calculated degree of coupling and the coupling threshold.

[0016] Compared with the prior art, the beneficial effects of the present invention are at least as follows: The technical solution provided in this application effectively reduces misjudgments caused by measurement noise and sampling errors by calculating the coupling degree between multiple data sequences and combining power data to eliminate interference during light load periods, thereby improving the reliability of anomaly detection. By monitoring the trend of coupling degree changes over time, gradual degradation of the equipment can be detected before the coupling degree falls below the threshold, enabling early warning of abnormal signs and avoiding sudden failures. The coupling threshold is dynamically calculated based on the mean and standard deviation of historical normal data, making the detection parameters adaptable to the electrical environment of different metering boxes and avoiding the applicability problem of fixed thresholds in different scenarios. The maximum coupling degree is calculated by shifting the data sequences along the time axis, eliminating the influence of sampling time offset or disturbance propagation delay between meters and improving the robustness of detection. By comparing the fluctuation of measurement data of each meter under light load, measurement loop anomalies are identified, and data points at abnormal times are eliminated, ensuring the accuracy of coupling degree calculation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of an embodiment of a method for detecting abnormal operation of a smart metering box in this application. Figure 2 This is a schematic diagram of one embodiment of an intelligent metering box operation anomaly detection system in this application. Detailed Implementation

[0019] This application provides a method and system for detecting abnormal operation of a smart metering box. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0020] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of a method for detecting abnormal operation of a smart metering box in this application includes: S1. Collect measurement data from each sub-meter connected to the same energy metering box at predetermined time intervals. For each sub-meter, construct a data sequence based on the measurement data acquired at multiple consecutive time intervals.

[0021] Specifically, under normal circumstances, multiple sub-meters (energy meters) connected to the same energy metering box are connected to the same power supply circuit. Therefore, the measurement data (voltage) they obtain should have a high degree of consistency. This consistency can be quantified by the degree of coupling. When all sub-meters (energy meters) connected to the same metering box are working normally, the degree of coupling between any two sub-meters is very close to 1. When any sub-meter malfunctions or its wiring is abnormal, the measurement data of the corresponding sub-meter will deviate from the true value. This deviation will cause the degree of coupling between the sub-meter and other normal sub-meters to decrease significantly. This application can identify abnormal sub-meters by calculating the degree of coupling between all sub-meters and based on the numerical distribution of the degree of coupling.

[0022] To identify anomalous sub-meters, measurement data is first collected at predetermined time intervals from each sub-meter connected to the same energy metering box. For each sub-meter, a data sequence is constructed based on the measurement data acquired over multiple consecutive time intervals, where the measurement data refers to voltage. This step clarifies the data acquisition object and the sequence construction method, providing standardized data input for subsequent coupling degree calculations.

[0023] S2. While collecting measurement data from each sub-meter, simultaneously collect power data from each sub-meter. Based on the power data, identify whether the power data of each sub-meter is less than the preset power threshold. If it is less than the power threshold, determine whether there is a measurement abnormality. If so, mark the measurement data at the corresponding collection time as invalid.

[0024] Specifically, when a sub-meter is under light load or no load, the signal amplitude measured by its internal sampling data (including voltage and frequency sampling) is small. At this time, the analog-to-digital converter operates near zero input, and the relative impact of quantization error and noise on the measurement data increases significantly. This causes random small fluctuations in the voltage reading. These random fluctuations are not caused by sub-meter malfunction but are normal measurement noise. Sub-meters with normal load do not have such fluctuations. However, these fluctuations reduce the coupling between the corresponding sub-meter and other normal meters, causing normally functioning sub-meters to be misjudged as abnormal. To avoid misjudgment, it is necessary to first determine whether the corresponding sub-meter has measurement abnormalities. If so, the measurement data during the light load period in the data sequence needs to be excluded. Therefore, while collecting the measurement data of each sub-meter, the power data of each sub-meter is collected simultaneously, and the power data is used to identify whether the power data of each sub-meter is less than a preset power threshold. If it is less than the power threshold, then... If any measurement anomalies are found in the corresponding sub-tables, the measurement data at the corresponding acquisition time is marked as invalid. In subsequent calculations of coupling degree, all data points acquired at the acquisition time corresponding to the marked invalid measurement data are removed from all data sequences. For example, if there are three sub-tables A, B, and C, and data was acquired at times 1, 2, 3, and 4 respectively: A1, A2, A3, A4, B1, B2, B3, B4, C1, C2, C3, C4, and A2 acquired at time 2 is marked as invalid, then, to ensure temporal alignment between data sequences, B2 and C2 acquired at time 2 are also removed when a measurement anomaly is determined at that time. This step, by removing data from potentially noisy periods, not only improves the accuracy of subsequent coupling degree calculations but also reduces false alarms caused by measurement noise, increasing the sensitivity for identifying truly abnormal devices.

[0025] S3. For any data sequence, calculate the degree of coupling with other data sequences, set a coupling threshold, and judge anomalies based on the calculated degree of coupling and the coupling threshold.

[0026] Specifically, to complete the core calculations for anomaly detection, the coupling degree between various data sequences is quantified. The coupling degree is typically expressed using the Pearson correlation coefficient, with values ​​ranging from -1 to 1, and normally close to 1. A coupling threshold is then set; the specific method for setting this threshold will be explained in detail later. This facilitates anomaly detection based on the calculated coupling degree and threshold. This step, by calculating the coupling degree and threshold, identifies single-shard anomalies, multi-shard anomalies, and overall anomalies, providing clear detection results for operations and maintenance personnel.

[0027] In one specific embodiment, anomalies are determined based on the calculated coupling degree and coupling threshold, specifically including the following steps: S31. If the coupling degree between all pairs of data sequences is greater than or equal to the coupling threshold, the metering box is judged to be operating normally. Otherwise, if there is a data sequence whose coupling degree with all other data sequences is less than the coupling threshold, the device corresponding to the data sequence is judged to be operating abnormally. S32. Otherwise, it indicates that there exists a set of data sequences consisting of at least two data sequences. Within the set of data sequences, the coupling degree between any two data sequences is less than the coupling threshold, and the coupling degree between any data sequence within the set and a data sequence outside the set is partially less than or partially greater than or equal to the coupling threshold. In this case, it is determined that there is a group anomaly.

[0028] Specifically, in order to determine the status of each sub-meter connected to the metering box, the coupling degree between each pair of data sequences is first calculated. Then, it is determined whether the coupling degree between each pair of all data sequences is greater than or equal to the coupling threshold. If so, it indicates that the measurement data (voltage) of all meters has a consistent trend and the metering box is operating normally. Otherwise, when the coupling degree between the data sequence of a certain sub-meter and the data sequences of all other sub-meters is less than the coupling threshold, it indicates that there is a problem with the sub-meter itself. Therefore, it is determined that the sub-meter corresponding to the data sequence is operating abnormally.

[0029] If neither of the above two conditions is met, it indicates the existence of a data sequence set consisting of at least two data sequences. Within this set, the coupling degree between any two data sequences is less than the coupling threshold, and the coupling degree between any data sequence within the set and a data sequence outside the set is partially less than or partially greater than or equal to the coupling threshold. Taking five partitions as an example, with the coupling threshold denoted by T, as shown in Table 1, the data sequence sets are 2 and 3. The coupling degree between data sequence 2 and data sequences 1 and 5 is greater than or equal to the coupling threshold, while the coupling degree between data sequence 2 and data sequences 3 and 4 is less than the coupling threshold. The coupling degree between data sequence 3 and data sequences 1 and 5 is greater than or equal to the coupling threshold, while the coupling degree between data sequence 3 and data sequences 2 and 4 is less than the coupling threshold. In this case, if 2 and 3 are excluded, the coupling degree between each pair of the remaining 1, 4, and 5 is greater than or equal to the preset threshold. Therefore, it is determined that there is a group anomaly, meaning that the partition corresponding to the data sequence within the data sequence set has an operational anomaly.

[0030] Table 1 In one specific embodiment, calculating the degree of coupling with other data sequences specifically includes the following steps: The current data sequence is referred to as the first data sequence. When calculating the coupling degree between the first data sequence and any other data sequence, the first data sequence is translated on the time axis relative to the other data sequences, and the coupling degree between the translated first data sequence and other data sequences is calculated.

[0031] Specifically, when collecting measurement data, each electricity meter may use an independent clock, and there may be time offsets between sampling times. In addition, when voltage harmonic disturbances occur in the power grid, the time when the disturbance reaches different meters may have slight differences. If the data at the original time is used directly to calculate the coupling degree, the time misalignment will reduce the coupling degree between two normal data sequences, which may lead to misjudgment. Therefore, when calculating the coupling degree between each data sequence and other data sequences, the current data sequence is shifted on the time axis. That is, the current data sequence is called the first data sequence. When calculating the coupling degree between the first data sequence and any other data sequence, the first data sequence is shifted relative to other data sequences on the time axis. The coupling degree between the shifted first data sequence and other data sequences is calculated. By shifting, the two data sequences can be aligned, thereby correctly reflecting their true correlation. Assume that the first data sequence is A, and its data points are arranged according to the sampling time as A(1), A(2), ..., A(n). The second data sequence is B, and the data points are B(1), B(2), ..., B(n). When calculating the coupling degree between A and B, A is shifted forward or backward along the time axis by several sampling intervals. For example, shifting A backward by k intervals yields a new sequence A', where A'(i) = A(i+k), and i represents the index of the data point in the shifted data sequence A. The coupling degree between A' and B is then calculated; the shift amount can be positive (backward) or negative (forward). The final coupling degree corresponds to the calculation result under a specific shift amount. The shift aims to reduce misjudgments caused by timing errors, making anomaly detection more reliable.

[0032] In one specific embodiment, calculating the coupling degree between the translated first data sequence and other data sequences includes: Other data sequences are referred to as the second data sequence. When the first data sequence is translated relative to the second data sequence on the time axis, multiple translations with different amounts are performed, and the coupling degree of the first and second data sequences under multiple translations is calculated to obtain multiple coupling degrees. The maximum coupling degree is taken as the final coupling degree of the first data sequence relative to the second data sequence.

[0033] Specifically, since the actual time offset is unknown, a single fixed translation may not precisely align the two sequences. Therefore, by trying multiple different translations to cover all possible offset ranges, the translation that maximizes the coupling is selected, ensuring the final result is not affected by improper translation selection. That is, the other data sequence is first referred to as the second data sequence. When translating the first data sequence relative to the second data sequence on the time axis, multiple translations with different amounts are performed, and the coupling degree between the first and second data sequences under these different translations is calculated. Multiple coupling degrees are obtained, and the maximum coupling degree is taken as the final coupling degree between the first and second data sequences. The offset corresponding to the maximum coupling degree is recorded. Subsequent calculations of the coupling degree between the first and second data sequences are based on this offset. This method, by using multiple different translations and selecting the maximum coupling degree as the final coupling degree, minimizes the impact of time delay or disturbance on the coupling degree calculation. This allows for accurate assessment of the true coupling degree between the two data sequences even with unknown time offsets, thereby improving the accuracy of anomaly detection.

[0034] In one specific embodiment, setting a coupling threshold specifically includes the following steps: Get the deployment time of the electricity metering box and each sub-meter. If the deployment time is less than or equal to the preset first time, set the coupling threshold to the first threshold. If the deployment duration is greater than or equal to the preset first duration, the coupling threshold is calculated based on historical data every preset time period.

[0035] Specifically, to improve the accuracy of anomaly detection by setting different coupling thresholds based on the operating time of the metering box, newly installed metering boxes and sub-meters lack sufficient historical operating data to statistically analyze the normal coupling degree distribution during the initial operation phase (i.e., when the deployment time is short). If calculation methods based on historical data are used at this time, the thresholds will be unreliable due to insufficient samples. Therefore, a pre-set empirical threshold, i.e., the first threshold, is used initially. Since the possibility of anomalies in the equipment is low during the initial operation phase, the first threshold can be set relatively high, such as 0.95, to ensure basic detection functions. After the system has been running for a sufficiently long time, such as when the deployment time is longer than the first threshold, a large amount of normal operation data has been accumulated. Therefore, the coupling threshold can be recalculated based on actual historical measurement data at each preset time period, such as monthly, so that the threshold better matches the actual operating characteristics of the energy metering box and improves the detection sensitivity. The above method can ensure that the newly installed metering box can operate normally and detect anomalies when data is insufficient, and can automatically optimize the coupling threshold according to the actual electrical environment during long-term operation, improving the detection sensitivity and adaptability, and reducing false alarms and missed alarms.

[0036] In one specific embodiment, calculating the coupling threshold based on historical data includes the following steps: Obtain the time period during which all devices are running normally in the recent period, calculate the degree of coupling between the data sequences of all sub-tables within the time period, calculate the mean and standard deviation of all coupling degrees, and set the coupling threshold as the mean minus a preset multiple multiplied by the standard deviation.

[0037] Specifically, different energy metering boxes have different electrical environments, meter accuracy, and load characteristics. A single fixed threshold may not be suitable for all scenarios. To provide a statistical and quantitative threshold calculation method that reflects the coupling degree distribution characteristics of the energy metering box under normal conditions and enables adaptive setting of the coupling threshold, this method analyzes measurement data under normal operating conditions, calculates the mean and standard deviation of the coupling threshold, and obtains a threshold that conforms to actual characteristics. First, a time period in which all sub-meters are operating normally is selected from historical data. The length of the time period can be set as needed, such as the past day. Within this time period, the coupling degree between every two different data sequences of all sub-meters is calculated to obtain multiple coupling degrees. Then, the mean and standard deviation of all coupling degrees are calculated, and a preset multiple is set. The preset multiple can be in the range of 3-6. The coupling threshold = mean - (preset multiple × standard deviation). Through the above method, the adaptive setting of the coupling threshold is achieved, so that the coupling threshold can dynamically match the actual operating characteristics of the metering box, improve the statistical rationality and accuracy of anomaly detection, and avoid misjudgment or missed judgment that may be caused by a fixed threshold in different environments.

[0038] In one specific embodiment, after calculating the coupling threshold, the following steps are also performed: Record the trend of the coupling degree between each data sequence and all other data sequences over time. When the coupling degree of a certain data sequence shows a monotonically decreasing trend over multiple consecutive sampling periods, and the current coupling degree is still higher than the coupling threshold, it is determined that the device corresponding to the data sequence has early abnormal signs.

[0039] Specifically, to detect early signs of gradual deterioration in equipment performance before the coupling degree drops below the coupling threshold, and to achieve predictive early warning and avoid sudden failures, some equipment failures are often gradual processes, such as capacitor aging in the sampling circuit and a slow increase in wiring contact resistance. These changes will cause the coupling degree of the data sequence of the energy meter with other energy meters to gradually decrease, but may still be higher than the set coupling threshold for a period of time. If an alarm is only triggered when the coupling degree is lower than the coupling threshold, there may already be significant metering errors or equipment failure. Therefore, by detecting the changing trend of the coupling degree, an early warning can be issued. During each anomaly detection, the coupling degree of each data sequence with other data sequences is calculated. For each data sequence, the calculated coupling degree and its corresponding time are recorded each time. Then, the coupling degree of the data sequence in the most recent consecutive sampling periods (such as 5 consecutive periods) is observed. If these values ​​show a monotonically decreasing trend of each time being lower than the previous one, and the latest coupling degree is still higher than the coupling threshold, then it is determined that the energy meter corresponding to the corresponding data sequence has early abnormal signs. The above methods can promptly detect the gradual deterioration trend of power supply, helping maintenance personnel to intervene before a fault occurs and improve power supply reliability.

[0040] In one specific embodiment, determining whether a measurement anomaly exists includes the following steps: For each sub-table, obtain its first measurement data corresponding to the collection time when all power data is less than the power threshold, calculate the fluctuation degree of the first measurement data, and at the same time calculate the fluctuation degree of the measurement data of other sub-tables at the same collection time. When the difference between the fluctuation degree of a certain sub-table and the average fluctuation degree of all other sub-tables is greater than the preset fluctuation threshold, it is determined that there is a measurement anomaly in the corresponding sub-table.

[0041] Specifically, when the sub-table is under light load or no load, the measured value may be affected by factors such as quantization error, environmental noise, and electromagnetic interference due to the small amplitude of the measured signal, resulting in certain fluctuations. If these fluctuations are too large, they may lead to inaccurate measurement data, which in turn leads to a reduction in the coupling degree. To avoid this situation, when the fluctuations are large, the data points at the corresponding time are removed. Therefore, by comparing the fluctuation levels of different sub-tables at the same time, data points that may cause interference are selectively removed, thereby accurately identifying abnormal sub-tables. First, for each acquisition time, record the power data of each sub-meter. When the power of a sub-meter is lower than a preset power threshold, use the measurement data collected at these times as the first measurement data. Calculate the fluctuation of the first measurement data, which can be represented by statistical measures such as variance, standard deviation, and range. Simultaneously, calculate the measurement data of other sub-meters at the corresponding acquisition time and calculate the fluctuation of these sub-meters at the corresponding time. Then, calculate the average fluctuation of all other sub-meters as the reference fluctuation level. By comparing the fluctuation of the target sub-meter with the reference fluctuation level, if the difference is greater than the preset fluctuation threshold, it is determined that the sub-meter has a measurement anomaly. Therefore, the data points at the corresponding acquisition time are removed from the data sequence when calculating the coupling degree in the subsequent calculation. If the difference is less than or equal to the fluctuation threshold, it indicates that the fluctuation of the sub-meter under light load is within the normal range. In this case, it is not necessary to remove the data sequence, and the original data sequence is used for calculation in the subsequent calculation of the coupling degree.

[0042] The above describes a method for detecting abnormal operation of a smart meter box according to an embodiment of this application. The following describes a system for detecting abnormal operation of a smart meter box according to an embodiment of this application. Please refer to [link / reference]. Figure 2 One embodiment of the intelligent metering box malfunction detection system in this application includes: The data acquisition module is used to collect measurement data from each sub-meter in the same energy metering box at predetermined time intervals. For each sub-meter, a data sequence is constructed based on the measurement data acquired at multiple consecutive time intervals. The data filtering module is used to collect power data of each sub-meter while collecting measurement data of each sub-meter. Based on the power data, it identifies whether the power data of each sub-meter is less than the preset power threshold. If it is less than the power threshold, the measurement data at the corresponding collection time is marked as invalid, and all data points collected at the collection time of invalid measurement data in all data sequences are removed. The anomaly detection module is used to calculate the degree of coupling between any given data sequence and other data sequences, set a coupling threshold, and determine anomalies based on the calculated degree of coupling and the coupling threshold.

[0043] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0044] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0045] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for detecting abnormal operation of an intelligent metering box, characterized in that, The method includes: S1. Collect measurement data from each sub-meter connected to the same energy metering box at predetermined time intervals. For each sub-meter, construct a data sequence based on the measurement data acquired at multiple consecutive time intervals. S2. While collecting the measurement data of each sub-meter, simultaneously collect the power data of each sub-meter. Based on the power data, identify whether the power data of each sub-meter is less than the preset power threshold. If it is less than the power threshold, determine whether there is a measurement abnormality. If so, mark the measurement data at the corresponding collection time as invalid. S3. For any data sequence, calculate the degree of coupling with other data sequences, set a coupling threshold, and judge anomalies based on the calculated degree of coupling and the coupling threshold.

2. The method according to claim 1, characterized in that, Anomalies are identified based on the calculated degree of coupling and coupling threshold, including: S31. If the coupling degree between all pairs of data sequences is greater than or equal to the coupling threshold, the metering box is judged to be operating normally. Otherwise, if there is a data sequence whose coupling degree with all other data sequences is less than the coupling threshold, the device corresponding to the data sequence is judged to be operating abnormally. S32. Otherwise, it indicates that there exists a set of data sequences consisting of at least two data sequences. Within the set of data sequences, the coupling degree between any two data sequences is less than the coupling threshold, and the coupling degree between any data sequence within the set and a data sequence outside the set is partially less than or partially greater than or equal to the coupling threshold. In this case, it is determined that there is a group anomaly.

3. The method according to claim 1, characterized in that, The degree of coupling between computation and other data sequences includes: The current data sequence is referred to as the first data sequence. When calculating the coupling degree between the first data sequence and any other data sequence, the first data sequence is translated on the time axis relative to the other data sequences, and the coupling degree between the translated first data sequence and other data sequences is calculated.

4. The method according to claim 3, characterized in that, Calculate the coupling degree between the translated first data sequence and other data sequences, including: Other data sequences are referred to as the second data sequence. When the first data sequence is translated relative to the second data sequence on the time axis, multiple translations with different amounts are performed, and the coupling degree of the first and second data sequences under multiple translations is calculated to obtain multiple coupling degrees. The maximum coupling degree is taken as the final coupling degree of the first data sequence relative to the second data sequence.

5. The method according to claim 1, characterized in that, Setting a coupling threshold includes: Get the deployment time of the electricity metering box and each sub-meter. If the deployment time is less than or equal to the preset first time, set the coupling threshold to the first threshold. If the deployment duration is greater than or equal to the preset first duration, the coupling threshold is calculated based on historical data every preset time period.

6. The method according to claim 5, characterized in that, The coupling threshold is calculated based on historical data, including: Obtain the time period during which all devices are running normally in the recent period, calculate the degree of coupling between the data sequences of all sub-tables within the time period, calculate the mean and standard deviation of all coupling degrees, and set the coupling threshold as the mean minus a preset multiple multiplied by the standard deviation.

7. The method according to claim 1, characterized in that, After calculating the coupling threshold, the following steps are also performed: Record the trend of the coupling degree between each data sequence and all other data sequences over time. When the coupling degree of a certain data sequence shows a monotonically decreasing trend over multiple consecutive sampling periods, and the current coupling degree is still higher than the coupling threshold, it is determined that the device corresponding to the data sequence has early abnormal signs.

8. The method according to claim 1, characterized in that, Determining whether there are measurement anomalies includes: For each sub-table, obtain the measurement data corresponding to all the collection times marked as invalid, calculate the fluctuation of the measurement data at the invalid times, and at the same time calculate the fluctuation of other sub-tables at the same collection times. When the difference between the fluctuation of a certain sub-table and the average fluctuation of all other sub-tables is greater than the preset fluctuation threshold, it is determined that there is a measurement anomaly in the corresponding sub-table.

9. A smart metering box malfunction detection system, used to implement the smart metering box malfunction detection method as described in any one of claims 1-8, characterized in that, The system includes: The data acquisition module is used to collect measurement data from each sub-meter in the same energy metering box at predetermined time intervals. For each sub-meter, a data sequence is constructed based on the measurement data acquired at multiple consecutive time intervals. The data filtering module is used to collect power data of each sub-meter while collecting measurement data of each sub-meter. Based on the power data, it identifies whether the power data of each sub-meter is less than the preset power threshold. If it is less than the power threshold, the measurement data at the corresponding collection time is marked as invalid, and all data points collected at the collection time of invalid measurement data in all data sequences are removed. The anomaly detection module is used to calculate the degree of coupling between any given data sequence and other data sequences, set a coupling threshold, and determine anomalies based on the calculated degree of coupling and the coupling threshold.

Images