Remote detection method and system for electric energy meter

Abstract

The present invention relates to the field of data processing, and in particular to a remote detection method and system for an electric energy meter. The method comprises: obtaining the operating data of the electric energy meter, using a variational mode decomposition algorithm to divide the operating data into windows according to a time series, obtaining a comprehensive frequency distribution curve of the operating data within the window, and generating a time-frequency graph; performing peak point detection based on the time-frequency graph, determining the main frequency peak point and the interference peak point, and calculating the degree of interference of the frequency of the electric energy meter data by harmonics based on the distribution characteristics between the main frequency peak point and the interference peak point. The change consistency and frequency amplitude of the interference degree value are used to calculate the abnormality degree value of the intelligent voltage meter and judge the abnormality of the electric energy meter. The present invention suppresses the interference of single-dimensional noise and improves the reliability of the abnormality judgment criterion by calculating the change consistency of data of different dimensions.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a remote detection method and system for an electric energy meter. Background Art

[0002] A low-voltage metering communication device is a device used in the low-voltage part of the power system. It is mainly used to collect, transmit and process data related to power metering. It usually combines metering and communication functions and is an important component of smart grid and Internet of Things technologies in the power system. For example, smart electricity meters can monitor the operating status of equipment in real time and detect abnormal conditions.

[0003] The existing Chinese patent application document with publication number CN117574192A discloses a method system, device, and storage medium for detecting energy meter errors. The method involves obtaining the topological connection of a substation; representing the power lines and user loads in the topological connection with lumped parameters to obtain an equivalent circuit for the substation; establishing a power balance model for the substation based on the equivalent circuit and the law of conservation of energy; collecting energy meter data over multiple metering periods; establishing a full-rank linear equation system using the collected energy meter data and the power balance model to obtain an energy meter remote error detection model; processing the coefficient matrix of the energy meter remote error detection model to obtain an energy meter correction coefficient; obtaining a relationship model between the energy meter error and the energy meter correction coefficient, and calculating an error detection result based on the energy meter correction coefficient and the relationship model.

[0004] The application document establishes an equivalent circuit model for the substation area and a power balance model, using collected data and correction coefficients to calculate errors to improve the accuracy and reliability of electricity metering. Currently, when calibrating electricity meters, it is necessary to consider the impact of harmonic interference on metering accuracy. However, traditional harmonic detection methods, primarily based on total harmonic distortion (THD), can provide overall information on harmonic content, but have difficulty distinguishing between transient interference and persistent anomalies, and lack quantitative analysis of frequency distribution similarity. This results in insufficient detection accuracy and cannot meet the requirements for high-precision detection of metering devices in smart grid and IoT environments. Summary of the Invention

[0005] To address the problem that harmonic interference makes it difficult for low-voltage metering devices to accurately distinguish between transient interference and persistent anomalies, and that they lack quantitative analysis of frequency distribution similarity, thus failing to meet the requirements of high-precision detection, the present invention provides solutions in the following aspects.

[0006] In a first aspect, a remote detection method for an electric energy meter comprises: obtaining operating data of the electric energy meter, wherein the operating data comprises voltage data and current data; dividing the operating data into windows according to a time series using a variational mode decomposition algorithm, obtaining a comprehensive frequency distribution curve of the operating data in the window, and generating a time-frequency diagram showing the change of the frequency of a reaction signal over time based on the comprehensive frequency distribution curve; performing peak point detection based on the time-frequency diagram to determine a main frequency peak point and an interference peak point, and calculating a value of the degree of interference of the frequency of the electric energy meter data by harmonics based on distribution characteristics between the main frequency peak point and the interference peak point, wherein the distribution characteristics comprise a relative interference degree and a deviation degree; calculating a value of the abnormality degree of the intelligent voltage meter based on the consistency of change of the interference degree value and the frequency amplitude, and judging the abnormality of the electric energy meter based on the abnormality degree value.

[0007] The effect is: through the variational mode decomposition (VMD) algorithm, the voltage and current data collected by the electricity meter are accurately analyzed, and the harmonic components in the signal are effectively separated. By constructing a time-frequency diagram, the change of signal frequency over time is monitored in real time, the peak points of the main frequency and the interference peak points are accurately identified, and the relative interference degree and deviation degree of the harmonics are calculated. This is conducive to providing an analytical basis for power quality analysis and anomaly detection. By analyzing the consistency of the change in the interference degree value and the frequency amplitude, the anomaly degree value is calculated, and a threshold is set to determine the abnormal state of the electricity meter, thereby optimizing maintenance plans, reducing maintenance costs, and improving the stability of the power system and user satisfaction. It not only meets the demand for high-precision detection of metering devices and significantly improves the accuracy and reliability of electricity metering, but also provides important technical support for the stable operation of the power system and power quality control.

[0008] Preferably, obtaining the comprehensive frequency distribution curve includes:

[0009] Each operating data is divided into windows at preset time intervals and into multiple data segments. The variational mode decomposition method is used to decompose the operating data in each window into multiple IMF components. Frequency analysis is performed on each IMF component to obtain the frequency distribution curve of the IMF component. The frequency distribution curves of all IMF components in the same window are superimposed to obtain the comprehensive frequency distribution curve of each window.

[0010] The effect is that through variational mode decomposition, different frequency components in the signal can be captured more finely. Compared with the traditional Fourier transform, variational mode decomposition provides an adaptive time-frequency analysis method without manual setting. This enables the algorithm to adapt to the complexity of the signal and effectively separate the various frequency components in the signal, including the fundamental frequency and various harmonics.

[0011] Preferably, obtaining the time-frequency graph includes:

[0012] The comprehensive frequency distribution curve of each window is used as the time node, the horizontal axis is the corresponding time point in the window, and the vertical axis is the frequency distribution at different time points to construct a time-frequency diagram.

[0013] The effect is that by analyzing the time-frequency diagram, the existence of harmonic interference and its evolution over time can be clearly identified, thereby evaluating the impact of harmonics on power quality, helping to detect abnormal conditions in the operation of electricity meters, such as transient interference or continuous anomalies, and providing data basis for fault diagnosis and maintenance.

[0014] Preferably, obtaining the interference level value includes:

[0015] Taking any moment on the time-frequency graph as the target moment, perform peak detection on the frequency curve of the window corresponding to the target moment to obtain a peak point sequence, take the peak point with the largest amplitude in the peak point sequence as the main frequency peak point, take the peak points other than the main frequency peak point as the interference peak point, calculate the absolute difference in amplitude between each interference peak point and the main frequency peak point, and take the ratio between the absolute difference and the main frequency peak point as the relative interference degree;

[0016] The relative interference degree is exponentially decayed using a negative exponential function and then summed up to obtain the interference degree value of each electric energy meter affected by harmonics at the target time.

[0017] Its effect is: by performing peak detection on the electricity meter data on the time-frequency graph, distinguishing the main frequency peak point and the interference peak point, and calculating the relative interference degree between them, and then using the negative exponential function to attenuate the interference degree and sum it up, the interference degree value of the electricity meter affected by harmonics at a specific moment is obtained, which can effectively quantify the impact of harmonic interference, improve the accuracy and reliability of electricity metering, and enhance the accuracy of power quality monitoring.

[0018] Preferably, obtaining the interference level value further includes:

[0019] Taking any moment on the time-frequency graph as the target moment, perform peak detection on the frequency curve of the window corresponding to the target moment to obtain a peak point sequence, take the peak point with the largest amplitude in the peak point sequence as the main frequency peak point, take the peak points other than the main frequency peak point as the interference peak point, calculate the absolute difference in amplitude between each interference peak point and the main frequency peak point, and take the ratio between the absolute difference and the main frequency peak point as the relative interference degree; calculate the ratio between the Gaussian function distribution of the local frequency curve corresponding to each interference peak point and the main frequency peak point Divergence value, the The difference between the divergence value and the relative interference degree is exponentially decayed using a negative exponential function and then summed to obtain the interference degree value of each electric energy meter affected by the harmonics at the target time.

[0020] Preferably, the The calculation of the divergence value includes:

[0021] Gaussian fitting is performed on the local frequency curve where the main frequency peak point is located and the local frequency curve where the interference peak point is located, and the local frequency distribution parameters of the main frequency peak point and the interference peak point are obtained respectively. The difference between the two Gaussian distributions is calculated based on the local frequency distribution parameters. Divergence value;

[0022] Preferably, the abnormality degree value satisfies the following relationship:

[0023] ;

[0024] Where, Indicates the The energy meter is The first dimension of the data The abnormality value of the peak point of the moment interference, Indicates the The energy meter is The first dimension of the data The consistency of change at all times, Indicates the The energy meter is The first dimension of the data The amount of change at a moment.

[0025] The effect is that by combining the consistency of changes and the amount of changes, the operating status of the meter can be monitored and evaluated in real time, and potential anomalies or failures can be discovered and prevented in a timely manner, thereby improving the accuracy of electricity metering and the reliability of the power system.

[0026] Preferably, judging the abnormality of the electric energy meter according to the abnormality degree value includes:

[0027] In response to the abnormality degree value of any dimension being greater than or equal to the abnormality threshold, it is determined that the operating state of the electric energy meter is abnormal. Conversely, if the abnormality degree values ​​of all dimensions are less than the abnormality threshold, it is determined that the operating state of the electric energy meter is normal.

[0028] Preferably, the method further comprises preprocessing the acquired operating data, wherein the preprocessing comprises removing outliers, filling missing data using interpolation, denoising the operating data, and performing normalization processing.

[0029] In a second aspect, a remote detection system for an electric energy meter includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned remote detection method for the electric energy meter is implemented.

[0030] The present invention has the following effects:

[0031] 1. The present invention uses the variational mode decomposition (VMD) algorithm to accurately separate the harmonic components in the electricity meter data. By analyzing the impact of harmonics on the main frequency, the interference level affected by the harmonics is identified, thereby avoiding the influence of harmonics on the electricity meter readings, which may lead to inaccurate detection of the operating status. This not only improves the accuracy of electricity metering, but also enhances the reliability of electricity data in the smart grid, meeting the demand for high-precision electricity metering in the smart grid and Internet of Things environments.

[0032] 2. The present invention utilizes The method of quantifying frequency distribution differences using divergence values ​​can accurately measure the Gaussian characteristic differences between the main frequency and the interference frequency, thereby improving the accuracy of harmonic interference detection. By analyzing the consistency of changes in data in different dimensions, it can effectively suppress single-dimensional noise interference and enhance the reliability of anomaly detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a method flow chart of steps S1 to S4 in a remote detection method for an electric energy meter according to an embodiment of the present invention.

[0034] Figure 2 The present invention is a block diagram of a remote detection system for an electric energy meter. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0036] Reference Figure 1 A remote detection method for an electric energy meter includes steps S1 to S4, specifically as follows:

[0037] S1: Acquire operating data of the electric energy meter, wherein the operating data includes voltage data and current data.

[0038] Exemplarily, the operating data of multiple smart energy meters are collected respectively, with a collection frequency of 512 times per second and a collection length of 10 seconds each time, so as to obtain the total operating data collected in the first 10 seconds of the current collection time for analysis.

[0039] The method also includes preprocessing the acquired operating data, wherein the preprocessing includes removing outliers, filling missing data using interpolation, denoising the operating data, and normalizing the operating data (unifying the amplitude range of the signal).

[0040] Statistical methods (such as the 3σ rule) or machine learning techniques are used to identify and remove these outliers. Missing data points are filled using interpolation. Common interpolation methods include linear interpolation, nearest neighbor interpolation, or model-based predictive interpolation. Digital filters (such as low-pass, high-pass, or band-pass filters) are applied to remove high-frequency noise from the data.

[0041] S2: Use the variational mode decomposition algorithm to divide the operating data into windows according to the time series, obtain the comprehensive frequency distribution curve of the operating data in the window, and generate a time-frequency diagram of the reaction signal frequency changing with time based on the comprehensive frequency distribution curve.

[0042] Each preset interval of operating data is divided into windows and multiple segments of data. The variational mode decomposition method is used to decompose the operating data in each window into multiple IMF components. Frequency analysis is performed on each IMF component to obtain the frequency distribution curve of the IMF component. The frequency distribution curves of all IMF components in the same window are superimposed to obtain the comprehensive frequency distribution curve of each window.

[0043] It should be noted that the preset interval is 0.5 seconds. Variational mode decomposition (VMD) is used for the operating data within each window. Due to its high computational complexity and the associated optimization issues, variational mode decomposition (VMD) is suitable for signals with multiple frequency components whose frequencies may vary over time. It can accurately separate the different frequency components in a signal and automatically determine the number of decomposed modes without manual configuration, making it suitable for complex signals.

[0044] A time-frequency graph is constructed using the comprehensive frequency distribution curve of each window as a time node, the horizontal axis as the corresponding time point within the window, and the vertical axis as the frequency distribution at different time points. A time-frequency graph can display the frequency components of a signal in different time periods.

[0045] In addition, the time-frequency diagram can be calculated by using the short-time Fourier transform (STFT) to obtain the time-frequency diagram. The short-time Fourier transform divides the signal into windows of fixed length, applies FFT to the data of each window, obtains the frequency distribution, and then constructs the time-frequency diagram with the frequency distribution of each window as the time node. The short-time Fourier transform has a fast calculation speed and is suitable for real-time or near real-time analysis, and is suitable for signals with relatively stable frequency components.

[0046] S3: Perform peak point detection based on the time-frequency graph to determine the main frequency peak point and the interference peak point. Calculate the degree of harmonic interference in the frequency of the electric energy meter data based on the distribution characteristics between the main frequency peak point and the interference peak point. The distribution characteristics include relative interference degree and deviation degree.

[0047] It should be noted that the main frequency peak point represents the normal operating frequency of the power system and is the reference point for power quality analysis. The interference peak point represents abnormal frequency components that may be caused by harmonics or other interference sources. These components may cause errors in power metering. The relative interference degree measures the impact of the interference peak relative to the main frequency peak. The deviation degree indicates the frequency deviation between the interference peak point and the main frequency peak point, reflecting the severity of the harmonic interference. The deviation degree analysis uses How the divergence value is calculated.

[0048] Obtaining the interference level value includes:

[0049] Taking any moment on the time-frequency graph as the target moment, perform peak detection on the frequency curve of the window corresponding to the target moment to obtain a peak point sequence. The peak point with the largest amplitude in the peak point sequence is taken as the main frequency peak point, and the peak points other than the main frequency peak point are taken as interference peak points. The absolute difference in amplitude between each interference peak point and the main frequency peak point is calculated, and the ratio between the absolute difference and the main frequency peak point is taken as the relative interference degree.

[0050] The relative interference degree is exponentially decayed using a negative exponential function and then summed up to obtain the interference degree value of each electric energy meter affected by harmonics at the target time.

[0051] Specifically, the interference level value satisfies the following relationship:

[0052] ;

[0053] Where, Indicates the The first The dimension data is in The interference degree value of the interference peak point at the moment to the main frequency peak point, Indicates the total number of interference peak points, Indicates the index of the interference peak point, Indicates the The dimension data is in Moment The absolute difference between the amplitude of the interference peak point and the main frequency peak point, Indicates the The dimension data is in The value of the maximum peak point at the moment, Expressed that The exponential function of the base.

[0054] That is to say, It means the The dimension data is in Moment The relative interference degree between the interference peak point and the main frequency peak point. The smaller the relative interference degree, the smaller the interference, because x represents a small response value, and thus the impact on the main frequency is lower. However, if the interference frequency is closer to the main frequency, the greater the impact on the main frequency, and the easier it is to deform the main frequency.

[0055] By identifying and quantifying harmonic interference, electric energy metering can be more accurately evaluated and corrected, thereby improving the accuracy of electric energy metering, helping to monitor and analyze the power quality of the power system, identify possible harmonic problems, and provide important technical support for the stable operation of smart grids and power quality control.

[0056] In addition, another embodiment further includes:

[0057] Taking any moment on the time-frequency graph as the target moment, perform peak detection on the frequency curve of the window corresponding to the target moment to obtain a peak point sequence. The peak point with the largest amplitude in the peak point sequence is taken as the main frequency peak point, and the peak points other than the main frequency peak point are taken as interference peak points. The absolute difference in amplitude between each interference peak point and the main frequency peak point is calculated, and the ratio between the absolute difference and the main frequency peak point is taken as the relative interference degree.

[0058] Calculate the Gaussian function distribution between each interference peak point and the local frequency curve corresponding to the main frequency peak point. Divergence values, respectively The divergence value and the relative interference degree are mapped using a negative exponential function. The mapped results are multiplied as the comprehensive weight, and the comprehensive weights of all interference peak points are added together to obtain the interference degree value of each electric energy meter affected by harmonics at the target time.

[0059] Specifically, the interference level value satisfies the following relationship:

[0060] ;

[0061] Where, Indicates the The first The dimension data is in The interference degree value of the moment interference peak point to the main frequency peak point, Indicates the total number of interference peak points, Indicates the index of the interference peak point, Indicates the The first smart meter The dimension data is in Moment The difference between the Gaussian function corresponding to the interference peak point and the main frequency peak point Divergence value, Indicates the Dimensional data in the Moment The absolute difference between the amplitude of the interference peak point and the main frequency peak point, Indicates the The value of the maximum peak point at the moment, Expressed that The exponential function of the base.

[0062] To further explain, different frequencies have different widths (which can be regarded as the discrete degree or distribution range of the frequency distribution), which means that the influence range of the frequency is different, and thus the The local frequency curve between the moment interference peak point and the two peak points on the left and right is calculated, and the mean variance of the local frequency curve is obtained to obtain the corresponding Gaussian function. Similarly, the local frequency curve between the maximum peak point and the two peak points on the left and right is obtained, and then the Gaussian function corresponding to the maximum peak point is obtained. If the two Gaussian functions are closer, then The smaller the divergence value, the greater the interference weight.

[0063] That is to say, It reflects the similarity of the local frequency distribution curves corresponding to the two peak points. The divergence value indicates that the two frequency distributions are highly similar, that is, The frequency distribution of the interference peak point is close to the frequency distribution of the main frequency peak point. The divergence value indicates that the two frequency distributions are quite different. The divergence value can be used to measure the impact of harmonic interference on the main frequency.

[0064] Get The divergence value includes the following steps:

[0065] Gaussian fitting is performed on the local frequency curve where the main frequency peak point is located and the local frequency curve where the interference peak point is located, and the local frequency distribution parameters of the main frequency peak point and the interference peak point are obtained respectively. The difference between the two Gaussian distributions is calculated based on the local frequency distribution parameters. Divergence value;

[0066] in, The divergence value satisfies the following relationship:

[0067] ;

[0068] Where, Indicates the The first The dimension data is in The moment The difference between the Gaussian function corresponding to the interference peak point and the main frequency peak point Divergence value, represents the natural logarithm function, Indicates the standard deviation of the Gaussian distribution of the peak point of the main frequency, Represents the standard deviation of the Gaussian distribution of the interference peak point, Represents the mean of the Gaussian distribution of the peak point of the main frequency, The mean of the Gaussian function distribution representing the peak point of interference.

[0069] That is, the local frequency distribution parameters are the mean and standard deviation, which describe the position and width of the Gaussian distribution, respectively, and reflect the central tendency and dispersion of the signal at a specific frequency.

[0070] That is, the smaller The divergence value means that the frequency distribution of the interference peak point is similar to that of the main frequency peak point, and the harmonic interference is small; on the contrary, a larger The divergence value means that the frequency distribution of the interference peak point is significantly different from the frequency distribution of the main frequency peak point, indicating that the harmonic interference is relatively large. The divergence value can be used to guide the optimization and adjustment of the power system to reduce harmonic interference and improve the accuracy of power metering and power quality.

[0071] Divergence and relative interference level are two metrics that measure interference from different perspectives. Divergence reflects the difference in frequency distribution shape between the interfering frequency band and the dominant frequency, while relative interference level reflects the difference in amplitude between the interfering frequency band and the dominant frequency. Combining these two metrics allows for a more comprehensive and accurate assessment of harmonic interference severity, taking into account both frequency distribution shape and amplitude differences.

[0072] S4: Calculate the abnormality level of the smart voltmeter based on the consistency of the change in the interference level value and the frequency amplitude, and determine the abnormality of the electric energy meter based on the abnormality level value.

[0073] Specifically, the abnormality degree satisfies the following relationship:

[0074] ;

[0075] Where, Indicates the The first The first dimension of the data The abnormality value of the peak point of the moment interference, Indicates the The energy meter is The first dimension of the data The consistency of change at all times, Indicates the The energy meter is The first dimension of the data The change in the interference level value at the moment.

[0076] That is to say, by analyzing data of different dimensions Moment and The absolute value of the difference in the harmonic interference degree at the moment is obtained. The variation of the interference degree value of the dimension data at two adjacent moments is calculated, and then the variance of the variation of the interference degree value of all two adjacent moments is calculated to obtain the first The consistency of change at all times.

[0077] Specifically, the change in the interference level value satisfies the following relationship:

[0078] ;

[0079] Where, Indicates the The first dimension of the data Moment and The change in the harmonic interference degree value at the moment, Indicates the The energy meter is The interference level value affected by harmonics at all times, Indicates the The energy meter is The interference degree value affected by harmonics at any moment.

[0080] That is to say, by calculating the absolute value of the difference, we can monitor the changes in single-dimensional data at consecutive time points. If the absolute value of the difference is abnormally large, it may indicate that some abnormal event or interference occurred between the two time points.

[0081] The Anomaly Score depends not only on the magnitude (amount) of change in the frequency distribution but also on the consistency of the change. If the change in the frequency distribution is consistent across multiple time points, then the change is likely normal or expected, and the Anomaly Score will be low. Conversely, if the change is inconsistent—that is, significant at some time points but little at others—this may indicate an anomaly and the Anomaly Score will be high.

[0082] In response to the abnormality degree value of any dimension being greater than or equal to the abnormality threshold, it is determined that the operating state of the electric energy meter is abnormal. Conversely, if the abnormality degree values ​​of all dimensions are less than the abnormality threshold, it is determined that the operating state of the electric energy meter is normal.

[0083] For example, the abnormal threshold is 1.5, and implementers can adjust it according to specific circumstances.

[0084] The present invention also provides a remote detection system for an electric energy meter. Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions. When the computer program instructions are executed by the processor, a remote detection method for an electric energy meter according to the first aspect of the present invention is implemented. The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are well known in the art and are therefore not described in detail here.

[0085] It should be noted that those skilled in the art may make various modifications and improvements without departing from the scope of the present invention, and these modifications and improvements fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be based on the appended claims.

Claims

1. A remote detection method for an electric energy meter, characterized in that: include: Acquiring operating data of the electric energy meter, wherein the operating data includes: voltage data and current data; The variational mode decomposition algorithm is used to divide the operating data into windows according to the time series, and the comprehensive frequency distribution curve of the operating data in the window is obtained. Based on the comprehensive frequency distribution curve, a time-frequency diagram of the reaction signal frequency changing with time is generated; Performing peak point detection based on the time-frequency graph to determine the main frequency peak point and the interference peak point, and calculating the degree of interference of the frequency of the electric energy meter data by the harmonics according to the distribution characteristics between the main frequency peak point and the interference peak point, wherein the distribution characteristics include: relative interference degree and deviation degree; Calculating the abnormality level of the smart voltmeter based on the consistency of the change in the interference level and the frequency amplitude, and determining the abnormality of the electric energy meter based on the abnormality level; Obtaining the interference degree value further includes: taking any moment on the time-frequency graph as the target moment, performing peak detection on the frequency curve of the window corresponding to the target moment to obtain a peak point sequence, taking the peak point with the largest amplitude in the peak point sequence as the main frequency peak point, taking the peak points other than the main frequency peak point as the interference peak point, calculating the absolute difference in amplitude between each interference peak point and the main frequency peak point, and taking the ratio between the absolute difference and the main frequency peak point as the relative interference degree; calculating the ratio between the Gaussian function distribution of the local frequency curve corresponding to each interference peak point and the main frequency peak point. Divergence values, respectively The divergence value and the relative interference degree are mapped using a negative exponential function. The mapped results are multiplied as the comprehensive weight. The comprehensive weights of all interference peak points are added together to obtain the interference degree value of each electric energy meter affected by harmonics at the target time. The abnormality degree value satisfies the following relationship: ； Where, Indicates the The first The first dimension of the data The abnormality value of the peak point of the moment interference, Indicates the The energy meter is The first dimension of the data The consistency of change at all times, Indicates the The energy meter is The first dimension of the data The change in the interference level value at the moment.

2. The remote detection method of an electric energy meter according to claim 1, characterized in that: Obtaining the comprehensive frequency distribution curve includes: Each operating data is divided into windows at preset time intervals and into multiple data segments. The variational mode decomposition method is used to decompose the operating data in each window into multiple IMF components. Frequency analysis is performed on each IMF component to obtain the frequency distribution curve of the IMF component. The frequency distribution curves of all IMF components in the same window are superimposed to obtain the comprehensive frequency distribution curve of each window.

3. The remote detection method of an electric energy meter according to claim 1, characterized in that: Acquiring the time-frequency graph includes: The comprehensive frequency distribution curve of each window is used as the time node, the horizontal axis is the corresponding time point in the window, and the vertical axis is the frequency distribution at different time points to construct a time-frequency diagram.

4. The remote detection method of an electric energy meter according to claim 1, characterized in that: Obtaining the interference level value includes: Taking any moment on the time-frequency graph as the target moment, perform peak detection on the frequency curve of the window corresponding to the target moment to obtain a peak point sequence, take the peak point with the largest amplitude in the peak point sequence as the main frequency peak point, take the peak points other than the main frequency peak point as the interference peak point, calculate the absolute difference in amplitude between each interference peak point and the main frequency peak point, and take the ratio between the absolute difference and the main frequency peak point as the relative interference degree; The relative interference degree is exponentially decayed using a negative exponential function and then summed up to obtain the interference degree value of each electric energy meter affected by harmonics at the target time.

5. The remote detection method of an electric energy meter according to claim 1, characterized in that: described The calculation of the divergence value includes: Gaussian fitting is performed on the local frequency curve where the main frequency peak point is located and the local frequency curve where the interference peak point is located to obtain the local frequency distribution parameters of the main frequency peak point and the interference peak point, and the difference between the two Gaussian distributions is calculated based on the local frequency distribution parameters. Divergence value.

6. The remote detection method of an electric energy meter according to claim 1, characterized in that: The determining of the abnormality of the electric energy meter according to the abnormality degree value includes: In response to the abnormality degree value of any dimension being greater than or equal to the abnormality threshold, it is determined that the operating state of the electric energy meter is abnormal. Conversely, if the abnormality degree values ​​of all dimensions are less than the abnormality threshold, it is determined that the operating state of the electric energy meter is normal.

7. The remote detection method of an electric energy meter according to claim 1, characterized in that: The method further includes preprocessing the acquired operating data, wherein the preprocessing includes removing outliers, filling missing data using an interpolation method, denoising the operating data, and normalizing the operating data.

8. A remote detection system for an electric energy meter, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the remote detection method of the electric energy meter according to any one of claims 1 to 7 is implemented.

Images