Abnormality detection method, device and equipment under smart metering correction and medium

Abstract

The application relates to the technical field of intelligent electric meters, and provides an abnormality detection method, device, equipment and medium under intelligent electric meter measurement correction, which can collect a non-single-dimensional data set, ensures the comprehensiveness of data relied on abnormality detection, calculates a current weight factor according to a steady-state working condition center vector and a distortion working condition center vector of dynamic calibration, can provide a dynamic weight adapting to a real working condition in the whole life cycle of the intelligent electric meter, processes pure waveform sequences and working condition environment time sequence data according to a multi-dimensional error model to obtain a current error, and respectively carries out independent measurement correction on a forward measurement channel and a reverse measurement channel of the intelligent electric meter according to the current error, so that accurate bidirectional measurement independent correction can be realized, and cross talk and word skipping problems can be avoided, and abnormality detection can be carried out according to the pure waveform sequences, forward electric quantity and reverse electric quantity, so that reliable abnormality detection can be realized based on high-precision measurement after bidirectional correction.

Description

Technical Field

[0001] This invention relates to the field of smart meter technology, and in particular to an anomaly detection method, device, equipment, and medium under smart meter metering correction. Background Technology

[0002] With the rapid development of smart grids and distributed energy, smart meters are facing increasingly complex grid conditions, and traditional anomaly detection technologies cannot meet the needs for accurate detection of abnormal electricity use behaviors such as electricity theft.

[0003] Specifically, due to the complexity of operating conditions, the metering accuracy of smart meters will continue to decrease with use. Therefore, if anomaly detection is based on inaccurate meter readings, it will lead to defects such as high false alarm rate, high missed detection rate, and inability to distinguish between electricity theft and equipment failure.

[0004] Therefore, there is an urgent need for a solution that can improve the reliability of anomaly detection in complex power grid environments based on metrological accuracy calibration. Summary of the Invention

[0005] In view of the above, it is necessary to provide an anomaly detection method, device, equipment and medium under smart meter metering correction, in order to solve the problem of low reliability of anomaly detection results in complex power grid environments.

[0006] An anomaly detection method under smart meter metering correction, the anomaly detection method under smart meter metering correction includes: In response to anomaly detection commands triggered by smart meters, multi-source time-series data is collected in real time. Obtain an electrical instantaneous value sequence including voltage and current parameters from the multi-source time-series data, and construct a working condition feature vector based on the electrical instantaneous value sequence; Obtain the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, and calculate the current weight factor based on the operating condition feature vector, the steady-state operating condition center vector, and the distorted operating condition center vector; The electrical instantaneous value sequence is subjected to low-pass filtering to obtain a first sequence, and the electrical instantaneous value sequence is subjected to continuous complex wavelet transform to obtain a second sequence; The first sequence and the second sequence are weighted and fused according to the current weight factor to obtain a pure waveform sequence; Obtain the constructed multidimensional error model, and obtain the working environment time series data from the multi-source time series data; The current error is obtained by processing the pure waveform sequence and the time series data of the working environment according to the multidimensional error model. Based on the current error, the forward metering channel and the reverse metering channel of the smart meter are independently metered and corrected to obtain the forward power and the reverse power. Anomaly detection is performed based on the pure waveform sequence, the positive charge, and the negative charge.

[0007] An anomaly detection device under smart meter metering correction, the smart meter metering correction anomaly detection device comprising: The data acquisition unit is used to collect multi-source time-series data in real time in response to anomaly detection commands triggered by smart meters. The construction unit is used to obtain an electrical instantaneous value sequence including voltage parameters and current parameters from the multi-source time-series data, and to construct a working condition feature vector based on the electrical instantaneous value sequence; The calculation unit is used to obtain the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, and to calculate the current weight factor based on the operating condition feature vector, the steady-state operating condition center vector and the distorted operating condition center vector. The processing unit is configured to perform low-pass filtering on the electrical instantaneous value sequence to obtain a first sequence, and to perform continuous complex wavelet transform on the electrical instantaneous value sequence to obtain a second sequence; A fusion unit is used to perform weighted fusion of the first sequence and the second sequence according to the current weighting factor to obtain a pure waveform sequence; The acquisition unit is used to acquire the constructed multidimensional error model and to acquire the working environment time series data from the multi-source time series data; The processing unit is further configured to process the pure waveform sequence and the operating environment time series data according to the multidimensional error model to obtain the current error; The correction unit is used to independently correct the forward metering channel and the reverse metering channel of the smart meter according to the current error, so as to obtain the forward power and the reverse power. An execution unit is configured to perform anomaly detection based on the pure waveform sequence, the positive charge, and the negative charge.

[0008] A computer device, the computer device comprising: A memory that stores at least one instruction; and a processor that executes the instructions stored in the memory to implement the anomaly detection method under the smart meter metering correction.

[0009] A computer-readable storage medium storing at least one instruction, which is executed by a processor in a computer device to implement the anomaly detection method under the metering correction of the smart meter.

[0010] As can be seen from the above technical solutions, this invention can collect multi-source time-series data in real time, providing a non-single-dimensional dataset and ensuring the comprehensiveness of the data on which anomaly detection depends. It calculates the current weighting factor based on the dynamically calibrated steady-state operating condition center vector and distorted operating condition center vector, providing dynamic weights adapted to real operating conditions throughout the entire lifecycle of the smart meter. It processes the clean waveform sequence and operating environment time-series data using a multi-dimensional error model to obtain the current error, and independently corrects the forward and reverse metering channels of the smart meter based on the current error, achieving accurate bidirectional metering correction and avoiding crosstalk and digit skipping issues. Anomaly detection is performed based on the clean waveform sequence, forward charge, and reverse charge, thus enabling reliable anomaly detection based on high-precision metering after bidirectional correction. Attached Figure Description

[0011] Figure 1 This is a flowchart of a preferred embodiment of the anomaly detection method under the smart meter metering correction of the present invention; Figure 2 This is a functional block diagram of a preferred embodiment of the anomaly detection device under the smart meter metering correction of the present invention; Figure 3 This is a schematic diagram of the structure of a computer device that implements a preferred embodiment of the anomaly detection method under smart meter metering correction according to the present invention. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0013] like Figure 1 The diagram shown is a flowchart of a preferred embodiment of the anomaly detection method under smart meter metering correction of the present invention. Depending on different requirements, the order of the steps in this flowchart can be changed, and some steps can be omitted.

[0014] The anomaly detection method under smart meter metering correction is applied to one or more computer devices. The computer device is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0015] The computer device can be any electronic product that can interact with the user, such as a personal computer, tablet computer, smartphone, personal digital assistant (PDA), interactive network television (IPTV), smart wearable device, etc.

[0016] The computer equipment may also include network equipment and / or user equipment. The network equipment includes, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of hosts or network servers.

[0017] The server can be a standalone server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0018] Artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0019] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0020] The network in which the computer device is located includes, but is not limited to, the Internet, wide area network, metropolitan area network, local area network, and virtual private network (VPN).

[0021] S10, in response to anomaly detection commands triggered by smart meters, collects multi-source time-series data in real time.

[0022] In this embodiment, the anomaly detection command can be triggered by maintenance personnel according to actual needs, or it can be triggered periodically according to a certain correction cycle. Subsequently, the data collected within a cycle can be processed.

[0023] In this embodiment, the multi-source timing data may include, but is not limited to: instantaneous voltage values, instantaneous current values, active power, reactive power, power factor, three-phase imbalance, the internal temperature of the smart meter, the power supply bias voltage, the ADC (Analog-to-Digital Converter) hardware gain coefficient, as well as the smart meter's cover opening signal, parameter change records, hardware error codes, etc.

[0024] For example, data acquisition can be performed using a dual-channel analog front-end (AFE) array, with each of the forward and reverse metering channels employing an independent AFE chain. This dual-channel analog front-end array may include manganin sampling resistors, high-speed analog electronic switches, a programmable gain amplifier (PGA), and a 24-bit analog-to-digital converter.

[0025] The forward metering channel and the reverse metering channel can be physically isolated by an isolation component to prevent bidirectional current signal crosstalk.

[0026] When the bidirectional current crosses zero, in order to avoid the interference of current spikes on the switching process, ensure the continuity and stability of sampling, and prevent switching noise from introducing measurement errors, nanosecond-level channel switching can be triggered, and forward and reverse physical isolation sampling can be performed.

[0027] S11, obtain an electrical instantaneous value sequence including voltage parameters and current parameters from the multi-source time series data, and construct a working condition feature vector based on the electrical instantaneous value sequence.

[0028] In this embodiment, constructing the operating condition feature vector based on the electrical instantaneous value sequence includes: The fundamental and harmonic components are extracted from the electrical instantaneous value sequence using Fast Fourier Transform, and the total harmonic distortion (THD) is calculated based on the fundamental and harmonic components. Extract each instantaneous current value and each instantaneous voltage value from the electrical instantaneous value sequence; The effective value of the current is calculated based on each instantaneous current value, and the ratio of the effective value of the current to the rated current is calculated as the load rate. Calculate the voltage variance based on each instantaneous voltage value; The rate of change of current at each moment is obtained by performing a difference operation on each two adjacent instantaneous current values, and the maximum rate of change is taken as the transient rate of change of current. The operating condition feature vector is constructed using the total harmonic distortion rate, the load factor, the voltage variance, and the current transient rate of change as elements.

[0029] The total harmonic distortion (THD) rate describes the degree to which voltage and current waveforms deviate from a standard sine wave, reflecting the intensity of harmonic interference in the power grid. The THD rate is the most direct and computationally cost-effective indicator for distinguishing between a clean power grid and a power grid affected by harmonic interference.

[0030] The load rate is used to reflect the current load level of the smart meter. It is simple to calculate and can effectively filter out misjudgments when the meter is unloaded or lightly loaded, thus improving the robustness of operating condition identification.

[0031] Among them, the voltage variance is used to quantify the degree of fluctuation of the grid voltage and judge the grid stability. It is the most intuitive quantitative indicator of grid stability.

[0032] The transient rate of change of current is used to capture sudden changes in current and identify transient behaviors such as load changes, electricity theft, and tampering.

[0033] This embodiment constructs the operating condition feature vector using the total harmonic distortion rate, the load rate, the voltage variance, and the current transient rate of change as elements. This provides multi-dimensional and comprehensive features, offering an effective data foundation for subsequent processing.

[0034] S12, obtain the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, and calculate the current weight factor based on the operating condition feature vector, the steady-state operating condition center vector and the distorted operating condition center vector.

[0035] In this embodiment, before obtaining the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, the method further includes: For each calibration cycle, the electrical timing data of the smart meter is traced backward from the current moment within a preset duration. Based on the element dimensions of the operating condition feature vector, construct the current steady-state feature vector and the current distortion feature vector according to the electrical timing data; Obtain the historical steady-state operating condition center vector and the historical distorted operating condition center vector of the previous calibration cycle; By fusing the historical steady-state operating condition center vector with the current steady-state feature vector according to the sliding forgetting factor, the steady-state operating condition center vector for each calibration period is obtained. By fusing the historical distortion condition center vector with the current distortion feature vector according to the sliding forgetting factor, the distortion condition center vector for each calibration period is obtained. Specifically, for the first calibration cycle, a clean steady-state power grid signal at the power frequency is input to the smart meter in a standard laboratory environment, and calibration is performed according to the element dimensions of the operating condition feature vector to obtain the initial steady-state operating condition center vector; distorted power grid signals of each harmonic are input to the smart meter in a standard laboratory environment, and calibration is performed according to the element dimensions of the operating condition feature vector to obtain the initial harmonic distortion operating condition center vector.

[0036] For example, in the first calibration cycle, a clean steady-state power grid signal and a distorted power grid signal mixed with specific harmonics can be input into the smart meter under a standard laboratory environment. Multiple sets of sample data are collected at standard temperatures, and the experimental optimal mean values ​​of total harmonic distortion rate, load factor, voltage variance, and current change rate are calculated as the initial steady-state operating condition center vector and the initial harmonic distortion operating condition center vector. Furthermore, during the actual grid-connected operation of the smart meter, historical time-series data can be automatically retrieved every 24 hours (i.e., one calibration cycle). When the power grid is determined to be in a standard no-load steady-state period or a high-frequency harmonic interference period, the current instantaneous feature vector is extracted, and corrected using an adaptive K-means recursive clustering center update algorithm with a sliding forgetting factor, resulting in the steady-state operating condition center vector and the distortion operating condition center vector for each calibration cycle.

[0037] The sliding forgetting factor can be a small value (e.g., 0.05) to prevent misjudgments of operating conditions caused by the long-term aging of components in smart meters or the drift of physical background noise due to distributed photovoltaic grid connection, thus ensuring the long-term accuracy of the weighting factors. Taking the steady-state operating condition center vector as an example, the steady-state operating condition center vector = (1 - sliding forgetting factor) × historical steady-state operating condition center vector + sliding forgetting factor × current steady-state feature vector.

[0038] In this embodiment, calculating the current weighting factor based on the operating condition feature vector, the steady-state operating condition center vector, and the distorted operating condition center vector includes: The first value is obtained by calculating the reciprocal of the square of the Euclidean distance between the characteristic vector of the operating condition and the center vector of the steady-state operating condition; The second value is obtained by calculating the reciprocal of the square of the Euclidean distance between the feature vector of the working condition and the center vector of the distorted working condition; Calculate the sum of the first value and the second value to obtain the third value; The quotient of the first value and the third value is calculated to obtain the current weight factor.

[0039] Under normal circumstances, the squared Euclidean distance between the working condition feature vector and the steady-state working condition center vector, and the squared Euclidean distance between the working condition feature vector and the steady-state working condition center vector will not be equal to 0. However, in order to deal with the extreme case, that is, the distance is 0 and the reciprocal calculation is meaningless, after calculating the squared Euclidean distance, each squared distance can be added to a minimum value to avoid the denominator being 0.

[0040] Specifically, the closer the operating condition feature vector is to the steady-state operating condition center vector, the larger the value of the current weight factor, and the higher the weight of the low-pass filter; the closer the operating condition feature vector is to the distorted operating condition center vector, the smaller the value of the current weight factor, and the higher the weight of the complex wavelet transform, thereby achieving adaptive waveform purification for different power grid operating conditions.

[0041] Through the above embodiments, the weighting ratio of subsequent low-pass filtering and complex wavelet transform can be dynamically adjusted to adapt to changes in the power grid and equipment after long-term operation.

[0042] S13, perform low-pass filtering on the electrical instantaneous value sequence to obtain a first sequence, and perform continuous complex wavelet transform on the electrical instantaneous value sequence to obtain a second sequence.

[0043] By performing low-pass filtering, high-frequency noise can be directly filtered out when the power grid is in a steady state and there is no large harmonic interference, leaving only the basic waveform.

[0044] By performing continuous complex wavelet transform processing, it is possible to accurately separate the entangled fundamental wave, harmonics, and interharmonics in harsh power grid environments (such as when there is a nearby photovoltaic grid connection or a large motor starting up, which can cause severe harmonics and waveform distortion). Conventional low-pass filtering will be distorted. Therefore, this embodiment uses complex wavelet transform to accurately separate the entangled fundamental wave, harmonics, and interharmonics.

[0045] S14, the first sequence and the second sequence are weighted and fused according to the current weight factor to obtain a pure waveform sequence.

[0046] In this embodiment, the first product of the current weight factor and the first sequence can be calculated, and the second product of (1 - current weight factor) and the second sequence can be calculated. The sum of the first product and the second product can be calculated to obtain the pure waveform sequence.

[0047] Among them, by dynamically calculating the weighting factor based on the current power consumption conditions, a current weighting factor close to 1 can be obtained under stable operating conditions with stable power consumption and no harmonics (when the operating condition feature vector is closer to the center vector of the stable operating condition, its distance term is smaller, infinitely approaching 0, and the first value of the penultimate term is larger, making the final calculated weighting factor closer to 1). Smart meters can prioritize the use of power-saving and fast low-pass filtering; under operating conditions with harmonic distortion in the power grid, a current weighting factor close to 0 can be obtained (when the operating condition feature vector is closer to the center vector of the stable operating condition, its distance term is smaller, infinitely approaching 0, and the first value of the penultimate term is larger, making the final calculated weighting factor closer to 1). The closer the quantity is to the center vector of the distortion condition, the smaller its distance term becomes, approaching 0 infinitely, while the second-to-last value of the term becomes larger, making the final calculated weight factor closer to 0. Smart meters can switch to high-precision complex wavelet transform across the board to forcibly extract the pure fundamental wave. During the soft switching process of the power grid gradually transitioning from steady state to distortion, the weight factor gradually slides from 1 to 0. Smart meters can be compatible with both types of filtering, avoiding the generation of step discontinuities (i.e., pseudo-noise) in the data at the moment of hard switching, ensuring that the data is smooth and faithful at all times.

[0048] The above embodiments can avoid noise caused by hard switching, output a clean waveform with a high signal-to-noise ratio, and provide high-quality input for subsequent measurement correction and anomaly detection.

[0049] S15, obtain the constructed multidimensional error model, and obtain the working environment time series data from the multi-source time series data.

[0050] In this embodiment, the multidimensional error model can be constructed based on five dimensions: temperature drift, voltage deviation, aging deviation, harmonic disturbance deviation, and inherent factory deviation.

[0051] In this embodiment, the operating environment time-series data may include the internal temperature of the smart meter, the power supply bias voltage, and the hardware gain coefficient.

[0052] The temperature inside the meter can be obtained by directly reading the temperature register value of the smart meter, or by sampling the voltage divider using an ADC and calculating the temperature value.

[0053] The power supply bias voltage is obtained by sampling the power rail voltage of the ADC.

[0054] The hardware gain coefficient is the factory calibration parameter of the smart meter.

[0055] S16, The current error is obtained by processing the pure waveform sequence and the time series data of the working environment according to the multidimensional error model.

[0056] In this embodiment, the process of processing the clean waveform sequence and the operating environment time series data according to the multidimensional error model to obtain the current error includes: The internal temperature of the smart meter is obtained from the time-series data of the operating environment, and the temperature drift is calculated based on the internal temperature. Obtain the power supply bias voltage from the operating environment time-series data, and calculate the voltage deviation based on the power supply bias voltage; Obtain the hardware gain coefficient from the timing data of the operating environment, and calculate the aging deviation based on the hardware gain coefficient; The current total harmonic distortion rate is calculated based on the pure waveform sequence, and the harmonic disturbance deviation is calculated based on the current total harmonic distortion rate. Obtain the inherent factory deviation of the smart meter; The current error is obtained by summing the temperature drift, voltage deviation, aging deviation, harmonic disturbance deviation, and inherent factory deviation.

[0057] Because the temperature drift of the smart meter chip is not linear, a quadratic curve shift will occur in the high and low temperature ranges. Therefore, this embodiment can use (a1×T+a2×T) 2 The nonlinear temperature drift is fitted to improve the fitting accuracy. Here, a1 and a2 are parameters, and T is the internal temperature of the smart meter.

[0058] Power supply voltage fluctuations directly affect the reference source and ADC reference voltage, leading to measurement deviations. Therefore, this embodiment uses b×ΔU to calculate the voltage deviation and compensate for measurement errors caused by power supply fluctuations. Here, b is a parameter, and ΔU is the power supply bias voltage.

[0059] The ADC gain drift (e.g., aging) decays slowly over time, more closely resembling a logarithmic curve than a linear change. Therefore, this embodiment can use c×ln(1+G) to fit this aging trend, which changes rapidly in the initial stage and stabilizes later, accurately modeling the logarithmic characteristics of device aging. Compared to the commonly used simple linear term, the logarithmic term provides better error compensation after long-term device operation. Here, c is a parameter, and G is the hardware gain coefficient.

[0060] Harmonics can cause deviations in the effective values ​​and power calculations of smart meters, and these deviations are approximately linearly related to the harmonic content. Therefore, this embodiment uses d×THD to calculate the harmonic disturbance deviation, which is used to correct the metering deviation caused by harmonic interference. Here, d is a parameter, and THD is the current total harmonic distortion rate.

[0061] Each smart meter has inherent deviations in its ADC, manganese copper resistors, and other components. This embodiment uses the inherent factory deviations to correct these basic errors in one go, which can be used to compensate for the factory fixed deviations.

[0062] Among them, a1, a2, b, c, and d are used as model parameters. The initial values ​​can be empirical values ​​determined based on historical data. They can be updated online through recursive least squares method, which solves the problem that traditional fixed models cannot adapt to device aging.

[0063] In the above embodiments, four major error sources—temperature, voltage, aging, and harmonics—are covered simultaneously, making it more applicable to a wider range of scenarios compared to a single temperature drift correction model.

[0064] S17, based on the current error, independently measure and correct the forward metering channel and the reverse metering channel of the smart meter to obtain the forward power and the reverse power.

[0065] In this embodiment, the step of independently measuring and correcting the forward and reverse metering channels of the smart meter based on the current error to obtain the forward and reverse electricity amounts includes: The current error is normalized to a decimal to obtain a normalized value; Calculate the difference between 1 and the normalized value to obtain the correction coefficient; Obtain the initial positive active power of the forward metering channel and the initial reverse active power of the reverse metering channel; The positive energy is obtained by multiplying the initial positive active energy by the correction coefficient; The reverse charge is obtained by multiplying the initial reverse active charge by the correction coefficient.

[0066] For example, when the current error is a percentage, the normalized value can be obtained by calculating the quotient of the current error and 100. Further, the difference between 1 and the normalized value can be used to obtain the correction coefficient. Even further, in the forward metering channel and the reverse metering channel, the product of the active power and the correction coefficient can be calculated respectively to obtain the corrected forward power and the reverse power.

[0067] The forward metering channel and the reverse metering channel are different hardware sampling channels with their own independent integration registers. Correction operations are applied to each channel separately, rather than a simple total charge correction. This effectively avoids mutual interference between forward and reverse charge levels, thus preventing crosstalk issues such as word skipping.

[0068] The original data is noisy and has high error, which may misjudge normal fluctuations as electricity theft. The data obtained after correction in this embodiment is pure, accurate and stable, with clear abnormal features, which can effectively reduce the false alarm rate of anomaly detection.

[0069] S18, perform anomaly detection based on the pure waveform sequence, the positive charge, and the negative charge.

[0070] When the electrical signal is abnormal, it is not necessarily caused by electricity theft. It may also be due to damage to the smart meter itself (such as hardware failure, or the parameter being tampered with by someone opening the cover). If this is not distinguished, there may be false alarms (such as mistaking equipment failure for electricity theft) or missed detections (such as mistaking electricity theft for equipment failure).

[0071] Therefore, in this embodiment, the anomaly detection based on the pure waveform sequence, the positive charge, and the reverse charge includes: Within each time window, a multi-dimensional feature matrix is ​​constructed based on the pure waveform sequence, including electrical characteristics, waveform distortion characteristics, timing fluctuation characteristics, and bidirectional electrical characteristics. Risk windows are selected based on the multidimensional feature matrix of each time window according to preset rules. The current window depth is calculated based on the current total harmonic distortion and the voltage variance; By comparing the number of risk windows with the current window depth, the temporal consistency verification result is obtained; When the timing consistency check result is a non-instantaneous disturbance, the total power is calculated based on the pure waveform sequence, and the power conservation deviation rate is calculated based on the total power. The energy conservation deviation rate is compared with the deviation rate threshold to obtain the energy conservation verification result. When the power conservation verification result is that the power is not conserved, the device status data of the smart meter is obtained; When the device status data is abnormal, it is determined that there is a risk of abnormal power consumption; or When the device status data is normal, it is determined that the smart meter has a hardware failure risk.

[0072] The electrical characteristics may include basic electrical features such as effective current value, power factor, three-phase unbalance, and voltage deviation rate.

[0073] The waveform distortion characteristics may include the proportion of each harmonic, phase offset, waveform distortion, etc.

[0074] The time-series fluctuation characteristics may include the power change rate of each time window, the window fluctuation variance, and the similarity of daily electricity consumption curves.

[0075] The bidirectional electrical characteristics may include the ratio of forward to reverse electrical quantities and the amount of reverse electrical quantity mutation.

[0076] The constructed multidimensional feature matrix can provide stable and highly discriminative features, thereby significantly reducing misjudgments in anomaly detection results.

[0077] The preset rules may include: (1) When the effective value of the current is close to 0, but the voltage is normal, it is marked as having a risk of undercurrent or short circuit for electricity theft; (2) When the proportion of harmonics exceeds 20%, it is marked as having a risk of harmonic injection for electricity theft; (3) When the power factor drops sharply to below 0.2 and it is a non-inductive load scenario, it is marked as having a risk of phase tampering; (4) When the three-phase imbalance exceeds the preset threshold, it is marked as a risk of three-phase load abnormality or wiring error; (5) When the reverse power fluctuation exceeds three times the historical average and there is no record of new energy access, it is marked as having the risk of reverse disguised discharge to steal electricity.

[0078] This embodiment first filters out high-risk windows so that multiple consecutive risk windows can be chained together for verification to determine whether the anomaly is a temporary interference or a continuous real anomaly. By eliminating transient interferences such as closing impact, motor start-stop, and instantaneous voltage fluctuations, the computational power consumption of subsequent complex verification can be reduced.

[0079] The basic window depth, harmonic influence coefficient, and voltage fluctuation influence coefficient can be read from the factory parameters of the smart meter.

[0080] The harmonic influence term is obtained by multiplying the harmonic influence coefficient by the current total harmonic distortion rate. The noisier the environment, the more instantaneous interference there is. In this case, the window depth will increase, requiring more consecutive window anomalies to eliminate interference and reduce the false alarm rate.

[0081] Specifically, the voltage fluctuation term is obtained by calculating the negative of the product of the voltage fluctuation influence coefficient and the voltage variance. A larger voltage variance indicates more severe fluctuations, suggesting poorer grid stability, but also making sustained stable anomalies less likely to occur. Therefore, when voltage fluctuates drastically, the window depth can be appropriately reduced to avoid missing genuine electricity theft attempts due to scattered anomalies caused by fluctuations.

[0082] The basic window depth is used to provide a minimum threshold for preventing false alarms, avoiding the judgment of anomalies based on only one window under any operating condition.

[0083] Furthermore, the sum of the harmonic influence term, the voltage fluctuation term, and the basic window depth can be calculated and rounded to the nearest integer to obtain the current window depth.

[0084] This embodiment calculates the dynamic window depth and adaptively adjusts the number of consecutive abnormal windows required based on the total harmonic distortion rate and voltage variance of the current power grid. Specifically, the higher the harmonic content, the larger the number of consecutive abnormal windows required to filter transient interference; the greater the voltage fluctuation, the smaller the number of consecutive abnormal windows required to avoid missing real anomalies. This mechanism achieves a dynamic balance between false alarm rate and response speed under different power grid operating conditions.

[0085] Specifically, when the number of risk windows is less than the current window depth, it indicates that there may be transient disturbances (such as closing impacts or motor starting), and the disturbance will not proceed to subsequent verification. When the number of risk windows is greater than or equal to the current window depth, it indicates that the disturbance is not transient and can proceed to subsequent verification. This can filter out transient fluctuations under normal operating conditions and significantly reduce the false alarm rate.

[0086] The theoretical total charge within a period can be calculated first based on the pure voltage and current instantaneous value sequence after dead-zone interpolation, using the composite Simpson integral method; the positive charge and the reverse charge after metering correction are added together to obtain the corrected total charge; the relative deviation between the corrected total charge and the theoretical total charge is calculated to obtain the charge conservation deviation rate.

[0087] When the energy conservation deviation rate is less than or equal to the deviation rate threshold (e.g., 2%), it indicates normal metering fluctuation and will not proceed to subsequent verification; when the energy conservation deviation rate is greater than the deviation rate threshold, it indicates that there is energy non-conservation and can proceed to subsequent verification, thereby enabling the identification of metering tampering and electricity theft that cannot be detected by the traditional single threshold method.

[0088] Specifically, if the device status data indicates the presence of a cover opening signal, parameter modification record, or hardware fault code, it means the meter has been opened or its parameters have been modified. There is a high probability that someone is using the cover opening to tamper with the meter and steal electricity, thus indicating a risk of abnormal electricity use. Conversely, if the device status data indicates the absence of a cover opening signal, parameter modification record, or hardware fault code, it means the meter has not been opened and its parameters have not been modified. This suggests a hardware malfunction (such as a damaged sampling module) rather than user electricity theft. In this case, it can be determined that the smart meter is experiencing a false alarm due to a hardware malfunction, thus preventing alarms from being triggered and eliminating interference from device malfunctions in anomaly detection, ensuring the accuracy of anomaly detection results.

[0089] In the above embodiments, the current window is first quickly judged based on the electrical, waveform, timing and bidirectional features in the multi-dimensional feature matrix to screen risk windows. Then, the risk windows are checked for consistency to filter instantaneous disturbances. Next, the power conservation is checked (such as cross-comparing the forward power, reverse power and the total power of waveform integration) to identify meter tampering behavior. Finally, the meter hardware status data is combined for joint verification to eliminate misjudgment of equipment failure. Thus, by combining behavioral features and physical tampering traces, anomaly detection is accurately performed, making the evidence chain of electricity theft more complete and more confident. At the same time, it can also avoid disputes and maintenance costs caused by misjudgment.

[0090] In this embodiment, after obtaining the anomaly detection result, an anomaly record can be generated synchronously, which may include timestamp, anomaly type, confidence level, comparison of data before and after correction, hash fingerprint, etc.

[0091] Furthermore, the abnormal records can be written into a one-time programmable memory for irreversible hardening. Moreover, the one-time programmable memory requires no additional erase / write operations, consumes less power, and does not require frequent writing, effectively extending device lifespan, reducing electronic waste from device disposal, and is more environmentally friendly.

[0092] The above embodiments enable abnormal data to be traceable and verifiable.

[0093] In this embodiment, after obtaining the anomaly detection result, the subsequent data processing strategy can be adjusted based on the detection result. For example, when it is determined that there is a risk of abnormal power consumption, the sampling rate of the smart meter can be increased to enhance the ability to capture high-frequency harmonics and transient interference. At the same time, the preprocessing intensity can be strengthened (such as increasing the order or cutoff frequency of the low-pass filter to enhance noise suppression; increasing the decomposition level of the complex wavelet transform to more finely separate harmonic components; improving the accuracy of dead-zone interpolation to reduce the metering error caused by the current dead zone, etc., so as to perform stronger purification processing on the high-interference waveforms collected under abnormal operating conditions and ensure the accuracy of subsequent metering and anomaly detection), and shorten the metering correction cycle. When it is determined that the steady state is normal, the sampling rate of the smart meter can be reduced, the correction cycle can be extended, and a low-power mode can be entered. When it is determined that there is a hardware failure risk of the smart meter, a meter replacement prompt can be issued.

[0094] In the above embodiments, when an abnormal power consumption risk is identified, the sampling rate of the smart meter is temporarily increased to enhance preprocessing. This ensures that additional computing power and energy consumption are only consumed when necessary, guaranteeing the accuracy of anomaly identification and metering. Furthermore, when a steady-state condition is determined to be normal, the sampling rate of the smart meter is reduced and the correction cycle is extended to enter a low-power mode, thereby significantly reducing the long-term average power consumption of the meter. This processing mechanism can dynamically adjust power consumption according to actual operating conditions, avoiding the ineffective energy consumption caused by a fixed high sampling rate. It achieves a dynamic balance between accuracy requirements and low-carbon energy saving, thus meeting the long-term development needs of low-carbon and environmentally friendly practices.

[0095] Furthermore, the aforementioned dynamic power consumption control mechanism significantly reduces the long-term average operating current and standby power consumption of smart meters, effectively reducing the average annual energy consumption of a single meter. Therefore, given the scale effect of the current massive number of smart meters, it can significantly reduce the indirect carbon emissions of the power system, meeting the requirements of green, low-carbon, energy-saving and loss-reducing, and possessing significant environmental benefits.

[0096] In this embodiment, the internal temperature of the smart meter can also be collected in real time, and online upgrades can be paused when the internal temperature exceeds the preset temperature range; if the internal temperature is within the safe range, firmware upgrades can be performed according to the normal procedure; if a temperature change occurs during the upgrade process, the firmware can be automatically rolled back in a safe mode to prevent device damage.

[0097] As can be seen from the above technical solutions, this invention can collect multi-source time-series data in real time, providing a non-single-dimensional dataset and ensuring the comprehensiveness of the data on which anomaly detection depends. It calculates the current weighting factor based on the dynamically calibrated steady-state operating condition center vector and distorted operating condition center vector, providing dynamic weights adapted to real operating conditions throughout the entire lifecycle of the smart meter. It processes the clean waveform sequence and operating environment time-series data using a multi-dimensional error model to obtain the current error, and independently corrects the forward and reverse metering channels of the smart meter based on the current error, achieving accurate bidirectional metering correction and avoiding crosstalk and digit skipping issues. Anomaly detection is performed based on the clean waveform sequence, forward charge, and reverse charge, thus enabling reliable anomaly detection based on high-precision metering after bidirectional correction.

[0098] like Figure 2The diagram shown is a functional block diagram of a preferred embodiment of the anomaly detection device under smart meter metering correction of the present invention. The anomaly detection device 11 under smart meter metering correction includes a data acquisition unit 110, a data construction unit 111, a calculation unit 112, a processing unit 113, a fusion unit 114, an acquisition unit 115, a correction unit 116, and an execution unit 117. The module / unit referred to in this invention refers to a series of computer program segments that can be executed by a processor and perform a fixed function, and are stored in memory. In this embodiment, the functions of each module / unit will be described in detail in subsequent embodiments.

[0099] The acquisition unit 110 is used to acquire multi-source time-series data in real time in response to an anomaly detection command triggered by a smart meter. The construction unit 111 is used to obtain an electrical instantaneous value sequence including voltage parameters and current parameters from the multi-source time series data, and to construct a working condition feature vector based on the electrical instantaneous value sequence; The calculation unit 112 is used to obtain the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, and to calculate the current weight factor based on the operating condition feature vector, the steady-state operating condition center vector and the distorted operating condition center vector. The processing unit 113 is used to perform low-pass filtering on the electrical instantaneous value sequence to obtain a first sequence, and to perform continuous complex wavelet transform on the electrical instantaneous value sequence to obtain a second sequence; The fusion unit 114 is used to perform weighted fusion of the first sequence and the second sequence according to the current weight factor to obtain a pure waveform sequence; The acquisition unit 115 is used to acquire the constructed multidimensional error model and to acquire the working environment time series data from the multi-source time series data. The processing unit 113 is further configured to process the pure waveform sequence and the working environment time series data according to the multidimensional error model to obtain the current error; The correction unit 116 is used to independently correct the forward metering channel and the reverse metering channel of the smart meter according to the current error, so as to obtain the forward power and the reverse power. The execution unit 117 is used to perform anomaly detection based on the pure waveform sequence, the positive charge, and the negative charge.

[0100] As can be seen from the above technical solutions, this invention can collect multi-source time-series data in real time, providing a non-single-dimensional dataset and ensuring the comprehensiveness of the data on which anomaly detection depends. It calculates the current weighting factor based on the dynamically calibrated steady-state operating condition center vector and distorted operating condition center vector, providing dynamic weights adapted to real operating conditions throughout the entire lifecycle of the smart meter. It processes the clean waveform sequence and operating environment time-series data using a multi-dimensional error model to obtain the current error, and independently corrects the forward and reverse metering channels of the smart meter based on the current error, achieving accurate bidirectional metering correction and avoiding crosstalk and digit skipping issues. Anomaly detection is performed based on the clean waveform sequence, forward charge, and reverse charge, thus enabling reliable anomaly detection based on high-precision metering after bidirectional correction.

[0101] like Figure 3 The diagram shown is a schematic representation of the computer device used in a preferred embodiment of the anomaly detection method for smart meter metering correction according to the present invention.

[0102] The computer device 1 may include a memory 12, a processor 13, and a bus (the arrow in the figure represents the bus), and may also include a computer program stored in the memory 12 and executable on the processor 13, such as an anomaly detection program under smart meter metering correction.

[0103] Those skilled in the art will understand that the schematic diagram is merely an example of computer device 1 and does not constitute a limitation on computer device 1. Computer device 1 can be either a bus topology or a star topology. Computer device 1 may also include more or fewer other hardware or software than shown in the diagram, or different component arrangements. For example, computer device 1 may also include input / output devices, network access devices, etc.

[0104] It should be noted that the computer device 1 described is merely an example. Other existing or future electronic products that are adaptable to this invention should also be included within the scope of protection of this invention and are incorporated herein by reference.

[0105] The memory 12 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 12 can be an internal storage unit of the computer device 1, such as a portable hard drive of the computer device 1. In other embodiments, the memory 12 can be an external storage device of the computer device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the computer device 1. Furthermore, the memory 12 can include both internal and external storage units of the computer device 1. The memory 12 can be used not only to store application software and various types of data installed on the computer device 1, such as the code of an anomaly detection program under smart meter metering correction, but also to temporarily store data that has been output or will be output.

[0106] In some embodiments, the processor 13 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 13 is the control unit of the computer device 1, connecting various components of the computer device 1 via various interfaces and lines. It executes programs or modules stored in the memory 12 (e.g., executing an anomaly detection program under smart meter metering correction), and calls data stored in the memory 12 to perform various functions of the computer device 1 and process data.

[0107] The processor 13 executes the operating system of the computer device 1 and various installed applications. The processor 13 executes these applications to implement the steps in the above embodiments of the anomaly detection method under smart meter metering correction, for example... Figure 1 The steps are shown.

[0108] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 12 and executed by the processor 13 to complete the present invention. The one or more modules / units may be a series of computer-readable instruction segments capable of performing specific functions, which describe the execution process of the computer program in the computer device 1. For example, the computer program may be divided into a collection unit 110, a construction unit 111, a calculation unit 112, a processing unit 113, a fusion unit 114, an acquisition unit 115, a correction unit 116, and an execution unit 117.

[0109] The integrated unit implemented as a software functional module described above can be stored in a computer-readable storage medium. This software functional module, stored in a storage medium, includes several instructions to cause a computer device (which may be a personal computer, computer equipment, or network device, etc.) or processor to execute portions of the anomaly detection method under smart meter metering correction described in the various embodiments of the present invention.

[0110] If the modules / units integrated in the computer device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware devices. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above.

[0111] The computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory, etc.

[0112] Furthermore, the computer-readable storage medium may primarily include a stored program area and a stored data area, wherein the stored program area may store the operating system, an application program required for at least one function, etc.; and the stored data area may store data created based on the use of blockchain nodes, etc.

[0113] The blockchain referred to in this invention is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0114] The bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, in... Figure 3 The bus is represented by only one straight line, but this does not mean that there is only one bus or one type of bus. The bus is configured to enable communication between the memory 12 and at least one processor 13, etc.

[0115] Although not shown, the computer device 1 may also include a power supply (such as a battery) to power various components. Preferably, the power supply can be logically connected to the at least one processor 13 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The computer device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0116] Furthermore, the computer device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the computer device 1 and other computer devices.

[0117] Optionally, the computer device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the computer device 1 and to display a visual user interface.

[0118] It should be understood that the embodiments described are for illustrative purposes only and are not limited to this structure in the scope of the patent application.

[0119] It will be understood by those skilled in the art that Figure 3 The structure shown does not constitute a limitation on the computer device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0120] Combination Figure 1 The memory 12 in the computer device 1 stores multiple instructions to implement an anomaly detection method under smart meter metering correction, and the processor 13 can execute the multiple instructions to achieve the following: In response to anomaly detection commands triggered by smart meters, multi-source time-series data is collected in real time. Obtain an electrical instantaneous value sequence including voltage and current parameters from the multi-source time-series data, and construct a working condition feature vector based on the electrical instantaneous value sequence; Obtain the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, and calculate the current weight factor based on the operating condition feature vector, the steady-state operating condition center vector, and the distorted operating condition center vector; The electrical instantaneous value sequence is subjected to low-pass filtering to obtain a first sequence, and the electrical instantaneous value sequence is subjected to continuous complex wavelet transform to obtain a second sequence; The first sequence and the second sequence are weighted and fused according to the current weight factor to obtain a pure waveform sequence; Obtain the constructed multidimensional error model, and obtain the working environment time series data from the multi-source time series data; The current error is obtained by processing the pure waveform sequence and the time series data of the working environment according to the multidimensional error model. Based on the current error, the forward metering channel and the reverse metering channel of the smart meter are independently metered and corrected to obtain the forward power and the reverse power. Anomaly detection is performed based on the pure waveform sequence, the positive charge, and the negative charge.

[0121] Specifically, the processor 13's implementation method for the above instructions can be found in [reference needed]. Figure 1 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0122] It should be noted that all the data involved in this case was legally obtained.

[0123] If any AI models, software tools, or components not belonging to this company appear in the embodiments of this invention, they are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this invention has been obtained by an entity authorized (with the knowledge and consent) or fully authorized by all parties through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0124] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0125] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0126] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0128] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0129] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0130] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices described in this invention can also be implemented by a single unit or device through software or hardware. Terms such as "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An anomaly detection method under smart meter metering correction, characterized in that, The anomaly detection method under smart meter metering correction includes: In response to anomaly detection commands triggered by smart meters, multi-source time-series data is collected in real time. Obtain an electrical instantaneous value sequence including voltage and current parameters from the multi-source time-series data, and construct a working condition feature vector based on the electrical instantaneous value sequence; Obtain the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, and calculate the current weight factor based on the operating condition feature vector, the steady-state operating condition center vector, and the distorted operating condition center vector; The electrical instantaneous value sequence is subjected to low-pass filtering to obtain a first sequence, and the electrical instantaneous value sequence is subjected to continuous complex wavelet transform to obtain a second sequence; The first sequence and the second sequence are weighted and fused according to the current weight factor to obtain a pure waveform sequence; Obtain the constructed multidimensional error model, and obtain the working environment time series data from the multi-source time series data; The current error is obtained by processing the pure waveform sequence and the time series data of the working environment according to the multidimensional error model. Based on the current error, the forward metering channel and the reverse metering channel of the smart meter are independently metered and corrected to obtain the forward power and the reverse power. Anomaly detection is performed based on the pure waveform sequence, the positive charge, and the negative charge.

2. The anomaly detection method under smart meter metering correction as described in claim 1, characterized in that, The step of constructing the operating condition feature vector based on the electrical instantaneous value sequence includes: The fundamental and harmonic components are extracted from the electrical instantaneous value sequence by fast Fourier transform, and the total harmonic distortion rate is calculated based on the fundamental and harmonic components. Extract each instantaneous current value and each instantaneous voltage value from the electrical instantaneous value sequence; The effective value of the current is calculated based on each instantaneous current value, and the ratio of the effective value of the current to the rated current is calculated as the load rate. Calculate the voltage variance based on each instantaneous voltage value; The rate of change of current at each moment is obtained by performing a difference operation on each two adjacent instantaneous current values, and the maximum rate of change is taken as the transient rate of change of current. The operating condition feature vector is constructed using the total harmonic distortion rate, the load factor, the voltage variance, and the current transient rate of change as elements.

3. The anomaly detection method under smart meter metering correction as described in claim 1, characterized in that, Before obtaining the dynamically calibrated steady-state operating condition center vector and the distorted operating condition center vector, the method further includes: For each calibration cycle, the electrical timing data of the smart meter is traced backward from the current moment within a preset duration. Based on the element dimensions of the operating condition feature vector, construct the current steady-state feature vector and the current distortion feature vector according to the electrical timing data; Obtain the historical steady-state operating condition center vector and the historical distorted operating condition center vector of the previous calibration cycle; By fusing the historical steady-state operating condition center vector with the current steady-state feature vector according to the sliding forgetting factor, the steady-state operating condition center vector for each calibration period is obtained. By fusing the historical distortion condition center vector with the current distortion feature vector according to the sliding forgetting factor, the distortion condition center vector for each calibration period is obtained. Specifically, for the first calibration cycle, a clean steady-state power grid signal at the power frequency is input to the smart meter in a standard laboratory environment, and calibration is performed according to the element dimensions of the operating condition feature vector to obtain the initial steady-state operating condition center vector; distorted power grid signals of each harmonic are input to the smart meter in a standard laboratory environment, and calibration is performed according to the element dimensions of the operating condition feature vector to obtain the initial harmonic distortion operating condition center vector.

4. The anomaly detection method under smart meter metering correction as described in claim 1, characterized in that, The step of calculating the current weighting factor based on the operating condition feature vector, the steady-state operating condition center vector, and the distorted operating condition center vector includes: The first value is obtained by calculating the reciprocal of the square of the Euclidean distance between the characteristic vector of the operating condition and the center vector of the steady-state operating condition; The second value is obtained by calculating the reciprocal of the square of the Euclidean distance between the feature vector of the working condition and the center vector of the distorted working condition; Calculate the sum of the first value and the second value to obtain the third value; The quotient of the first value and the third value is calculated to obtain the current weight factor.

5. The anomaly detection method under smart meter metering correction as described in claim 2, characterized in that, The current error is obtained by processing the pure waveform sequence and the operating environment time series data according to the multidimensional error model, including: The internal temperature of the smart meter is obtained from the time-series data of the operating environment, and the temperature drift is calculated based on the internal temperature. Obtain the power supply bias voltage from the operating environment time-series data, and calculate the voltage deviation based on the power supply bias voltage; Obtain the hardware gain coefficient from the timing data of the operating environment, and calculate the aging deviation based on the hardware gain coefficient; The current total harmonic distortion rate is calculated based on the pure waveform sequence, and the harmonic disturbance deviation is calculated based on the current total harmonic distortion rate. Obtain the inherent factory deviation of the smart meter; The current error is obtained by summing the temperature drift, voltage deviation, aging deviation, harmonic disturbance deviation, and inherent factory deviation.

6. The anomaly detection method under smart meter metering correction as described in claim 1, characterized in that, The step of independently correcting the forward and reverse metering channels of the smart meter based on the current error to obtain the forward and reverse electricity amounts includes: The current error is normalized to a decimal to obtain a normalized value; Calculate the difference between 1 and the normalized value to obtain the correction coefficient; Obtain the initial positive active power of the forward metering channel and the initial reverse active power of the reverse metering channel; The positive energy is obtained by multiplying the initial positive active energy by the correction coefficient; The reverse charge is obtained by multiplying the initial reverse active charge by the correction coefficient.

7. The anomaly detection method under smart meter metering correction as described in claim 5, characterized in that, The anomaly detection based on the pure waveform sequence, the positive charge, and the negative charge includes: Within each time window, a multi-dimensional feature matrix is ​​constructed based on the pure waveform sequence, including electrical characteristics, waveform distortion characteristics, timing fluctuation characteristics, and bidirectional electrical characteristics. Risk windows are selected based on the multidimensional feature matrix of each time window according to preset rules. The current window depth is calculated based on the current total harmonic distortion and the voltage variance; By comparing the number of risk windows with the current window depth, the temporal consistency verification result is obtained; When the timing consistency check result is a non-instantaneous disturbance, the total power is calculated based on the pure waveform sequence, and the power conservation deviation rate is calculated based on the total power. The energy conservation deviation rate is compared with the deviation rate threshold to obtain the energy conservation verification result. When the power conservation verification result is that the power is not conserved, the device status data of the smart meter is obtained; When the device status data is abnormal, it is determined that there is a risk of abnormal power consumption; or When the device status data is normal, it is determined that the smart meter has a hardware failure risk.

8. An anomaly detection device under smart meter metering correction, characterized in that, Used to perform the anomaly detection method under smart meter metering correction as described in any one of claims 1 to 7.

9. A computer device, characterized in that, The computer device includes: A memory that stores at least one instruction; and a processor that executes the instructions stored in the memory to implement the anomaly detection method under smart meter metering correction as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one instruction, which is executed by a processor in a computer device to implement the anomaly detection method under smart meter metering correction as described in any one of claims 1 to 7.

Images