Electric energy meter remote monitoring and fault diagnosis system based on Internet of Things

Abstract

The invention provides an electric energy meter remote monitoring and fault diagnosis system based on the Internet of Things, and belongs to the technical field of electrical quantity measurement, and the system comprises a monitoring module which collects real-time electric energy data in real time, and obtains historical electric energy data; the analysis module is used for determining a historical feature vector of each electric energy meter; the trend module is used for determining each statistical characteristic of each electric energy meter and a trend value of each regular characteristic, and determining a periodic trend vector of each electric energy meter; the fault module is used for determining real-time feature vectors, performing fault diagnosis on all the electric energy meters of the monitoring terminal and determining fault electric energy data of each electric energy meter; and the interaction module is used for displaying the corresponding fault electric energy data on each electric energy meter at the monitoring terminal and giving an alarm. The data comprehensiveness can be improved, the dynamic characteristics of electric energy meter operation can be comprehensively captured, the characteristic change rule can be accurately reflected, the fault or trend change identification capability can be improved, and the accuracy, efficiency and adaptability of electric energy meter fault diagnosis can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical quantity measurement, and particularly to a remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things. Background Art

[0002] The fault diagnosis technology of electric energy meters originated from the development of the electric power industry in the early 20th century. Initially, it relied on manual inspections and simple instrument detections, such as abnormal voltage measurements by voltmeters. With the progress of electronic technology, in the 1970s, automatic monitoring devices based on fixed thresholds emerged, which could detect situations where the voltage or current exceeded the set range, but had limited recognition of complex faults (such as intermittent abnormalities). Entering the 21st century, with the popularization of smart meters, combined with microprocessor and communication technologies, remote data transmission and basic fault alarms were realized, such as the AMI (Advanced Metering Infrastructure) system in the 2000s. However, most of these technologies rely on static rules and lack dynamic trend analysis, making it difficult to adapt to diverse electricity consumption patterns. In recent years, with the rise of the Internet of Things technology, it has been gradually applied to electric energy monitoring since 2010, introducing sensor networks and cloud processing, which have improved the data acquisition frequency and coverage. Nevertheless, existing systems still mainly rely on single-feature judgment, with insufficient utilization of historical data, and there is still room for improvement in the accuracy and real-time performance of fault diagnosis.

[0003] Therefore, the present invention provides a remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things. Summary of the Invention

[0004] The present invention provides a remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things. By analyzing the obtained historical electric energy data, the historical feature vectors of each electric energy meter, all statistical features, the trend values of all regular features, and the periodic trend vectors are determined. According to the real-time electric energy data and the periodic trend vectors of all electric energy meters, the fault electric energy data of each electric energy meter is determined, and the fault electric energy data of each electric energy meter is displayed on the electric energy meter and an alarm is issued, which can improve the comprehensiveness of data, comprehensively capture the dynamic characteristics of the operation of the electric energy meter, accurately reflect the change law of features, improve the recognition ability of faults or trend changes, and improve the accuracy, efficiency, and adaptability of electric energy meter fault diagnosis.

[0005] The present invention provides a remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things, including: Monitoring module: Real-time collect the real-time electric energy data of all electric energy meters of the monitoring terminal, and obtain the historical electric energy data of all electric energy meters of the monitoring terminal; Analysis module: Analyze the historical electric energy data, determine the periodic statistical vector and the periodic regular vector of each electric energy meter of the monitoring terminal, and determine the historical feature vector of each electric energy meter; Trend module: Based on the periodic statistical vector and the periodic law vector, determine the trend values of each statistical feature and each law feature of each electricity meter, and determine the periodic trend vector of each electricity meter; Fault module: Determine the real-time feature vector based on the real-time electricity data. Based on the real-time feature vector and the periodic trend vectors of all electricity meters, perform fault diagnosis on all electricity meters of the monitoring terminal to determine the fault electricity data of each electricity meter; Interaction module: Display the corresponding fault electricity data for each electricity meter of the monitoring terminal and issue an alarm.

[0006] According to the remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by the present invention, the monitoring module includes: Monitoring device group unit: Based on the distribution location of the electricity meters of the monitoring terminal and the monitoring requirements, determine the data monitoring device group and the installation location of each data monitoring device in the data monitoring device group; Real-time electricity sub-data unit: Based on the installation locations of all data monitoring devices in the data monitoring device group, install the data monitoring device group. Based on the installed data monitoring device group, collect the real-time electricity sub-data of each electricity meter of the monitoring terminal in the real-time monitoring period. The real-time electricity sub-data at least includes real-time voltage data, real-time power data, and real-time electricity consumption data; Real-time electricity data unit: Preprocess the real-time electricity sub-data of each electricity meter of the monitoring terminal. Based on the preprocessed real-time electricity sub-data of all electricity meters of the monitoring terminal, determine the real-time electricity data; Historical electricity sub-data unit: Extract the historical electricity sub-data of each electricity meter of the monitoring terminal in multiple historical monitoring periods from the database. Among them, the historical electricity sub-data includes multiple historical period data. Among them, the historical period data includes periodic voltage data, periodic power data, and periodic electricity consumption data; Historical electricity data unit: Based on the extracted historical electricity sub-data of all electricity meters of the monitoring terminal, determine the historical electricity data.

[0007] According to the remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by the present invention, the analysis module includes: Periodic statistical vector unit: Extract statistical features from each historical period data in the historical electricity sub-data of each electricity meter in the historical electricity data, and determine the periodic statistical vector of each electricity meter in each historical monitoring period; Voltage-power double broken line graph unit: Based on the periodic voltage data and periodic power data in each historical period data in the historical electricity sub-data of each electricity meter in the historical electricity data, draw the voltage-power double broken line graph of each electricity meter in each historical monitoring period; Electricity quantity discount graph unit: Based on the periodic electricity quantity data in each historical cycle data of the historical electricity sub-data of each electricity meter in the historical electricity data, draw the electricity quantity discount graph of each electricity meter in each historical monitoring cycle; Periodic law vector unit: Extract the law characteristics from the voltage-power double discount graph and the electricity quantity discount graph of each electricity meter in each historical monitoring cycle, and determine the periodic law vector of each electricity meter in each historical monitoring cycle; Historical feature vector unit: Based on the periodic statistical vector and the periodic law vector of each electricity meter, determine the historical feature vector of each electricity meter.

[0008] According to the remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by the present invention, the trend module includes: Trend unit: Analyze the periodic statistical vector and the periodic law vector of each electricity meter in all historical monitoring cycles, and determine the trend value of each statistical feature in the periodic statistical vector and the trend value of each law feature in the periodic law vector of each electricity meter; Periodic trend vector unit: Based on the trend values of all statistical features in the periodic statistical vector and the trend values of all law features in the periodic law vector of each electricity meter, determine the periodic trend vector of each electricity meter.

[0009] According to the remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by the present invention, the trend value unit includes: First calculation unit: Based on the periodic statistical vector of each electricity meter in all historical monitoring cycles, calculate the trend value of each statistical feature in the periodic statistical vector of each electricity meter; ; ; ; ; Wherein, represents the trend value of the a-th statistical feature in the periodic statistical vector of the i-th electricity meter, represents the change trend of the a-th statistical feature in the periodic statistical vectors of all historical monitoring cycles of the i-th electricity meter, represents the sign of the change trend of the a-th statistical feature in the periodic statistical vectors of all historical monitoring cycles of the i-th electricity meter, represents the trend reliability value of the a-th statistical feature in the periodic statistical vectors of all historical monitoring cycles of the i-th electricity meter, represents the first trend reliability threshold, represents the second trend reliability threshold, denotes the average feature change value of the ath statistical feature in the periodic statistical vectors of all historical monitoring periods of the ith electricity meter, and N1 denotes the number of extracted historical monitoring periods. denotes the eigenvalue of the ath statistical feature in the periodic statistical vector of the jth historical monitoring period of the ith electricity meter. denotes the eigenvalue of the ath statistical feature in the periodic statistical vector of the (j + 1)th historical monitoring period of the ith electricity meter. denotes the average eigenvalue of the ath statistical feature in the periodic statistical vectors of all historical monitoring periods of the ith electricity meter. denotes the monitoring time of the jth historical monitoring period. denotes the average monitoring time of all historical monitoring periods. The second calculation unit: calculates the trend value of each regularity feature in the periodic regularity vector of each electricity meter based on the periodic regularity vectors of each electricity meter in each historical monitoring period. ; ; ; ; denotes the change trend of the bth regularity feature in the periodic regularity vector of the ith electricity meter. denotes the change trend of the bth regularity feature in the periodic regularity vectors of all historical monitoring periods of the ith electricity meter. denotes the sign of the change trend of the bth regularity feature in the periodic regularity vectors of all historical monitoring periods of the ith electricity meter. denotes the trend reliability value of the bth regularity feature in the periodic regularity vectors of all historical monitoring periods of the ith electricity meter. denotes the average feature change value of the bth regularity feature in the periodic regularity vectors of all historical monitoring periods of the ith electricity meter. denotes the eigenvalue of the bth regularity feature in the periodic regularity vector of the jth historical monitoring period of the ith electricity meter. denotes the eigenvalue of the bth regularity feature in the periodic regularity vector of the (j + 1)th historical monitoring period of the ith electricity meter. denotes the average eigenvalue of the bth regularity feature in the periodic regularity vectors of all historical monitoring periods of the ith electricity meter.

[0010] According to the remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by the present invention, the fault module includes: Real-time statistical vector unit: Extract statistical features from the real-time power sub-data of each electricity meter in the real-time power data, and determine the real-time statistical vector of each electricity meter within the real-time monitoring period; Line graph unit: Based on the real-time voltage data and real-time power data in the real-time power sub-data of each electricity meter in the real-time power data, draw a real-time first double line graph. At the same time, based on the real-time power consumption data in the real-time power sub-data of each electricity meter in the real-time power data, draw a real-time second line graph; Real-time pattern vector unit: Extract pattern features from the real-time first line graph and real-time second line graph of each electricity meter, and determine the real-time pattern vector of each electricity meter within the real-time monitoring period; Real-time feature vector unit: Based on the real-time statistical vector and real-time pattern vector of each electricity meter within the real-time monitoring period, determine the real-time feature vector.

[0011] According to the remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by the present invention, the fault module further includes: Eigenvalue range unit: Based on the historical feature vectors and periodic trend vectors of each electricity meter in the historical monitoring period closest to the real-time monitoring period, determine the eigenvalue range of each statistical feature and each pattern feature of each electricity meter in the real-time monitoring period; Judgment unit: Judge whether each statistical feature and each pattern feature in the real-time feature vector of each electricity meter within the real-time monitoring period are within the corresponding eigenvalue ranges; Fault feature vector unit: For any statistical feature or pattern feature that is not within the eigenvalue range, based on all the statistical features and all the pattern features of each electricity meter that are not within the eigenvalue range, determine the fault feature vector of each electricity meter. Otherwise, determine that the fault feature vector of the electricity meter is empty; Fault power data unit: Based on the fault feature vector of each electricity meter, determine the fault power data of each electricity meter.

[0012] According to the remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by the present invention, the eigenvalue range unit includes: ; ; ; ; Wherein, represents the eigenvalue range of the a-th statistical feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter, Denote the predicted feature value of the a-th statistical feature in the periodic statistical vector of the real-time monitoring period of the i-th electricity meter. Denote the standard deviation of the feature values of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter. Denote the feature value of the a-th statistical feature in the periodic statistical vector of the N1-th historical monitoring period of the i-th electricity meter. Denote the monitoring time of the real-time monitoring period. Denote the monitoring time of the N1-th historical monitoring period. Denote the range of the feature values of the b-th regular feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter. Denote the predicted feature value of the b-th regular feature in the periodic regularity vector of the real-time monitoring period of the i-th electricity meter. Denote the feature value of the b-th regular feature in the periodic regularity vector of the N1-th historical monitoring period of the i-th electricity meter. Denote the standard deviation of the feature values of the b-th regular feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter.

[0013] Compared with the prior art, the beneficial effects of the present application are as follows: By analyzing the obtained historical electricity data, determine the historical feature vectors, all statistical features, and trend values of all regular features, as well as the periodic trend vectors of each electricity meter. According to the real-time electricity data and the periodic trend vectors of all electricity meters, determine the faulty electricity data of each electricity meter, display the faulty electricity data of each electricity meter on the electricity meter and issue an alarm, which can improve the comprehensiveness of data, comprehensively capture the dynamic characteristics of the operation of the electricity meter, accurately reflect the change law of features, improve the recognition ability of faults or trend changes, and improve the accuracy, efficiency, and adaptability of the fault diagnosis of the electricity meter. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0015] Figure 1 It is a schematic structural diagram of a remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] To make the objectives, technical solutions and advantages of the present invention more clear, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0017] Embodiment 1: The embodiment of the present invention provides a remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things, as Figure 1 shown, including: Monitoring module: Real-time collect the real-time electric energy data of all electric energy meters of the monitoring terminal, and obtain the historical electric energy data of all electric energy meters of the monitoring terminal; Analysis module: Analyze the historical electric energy data, determine the periodic statistical vector and periodic law vector of each electric energy meter of the monitoring terminal, and determine the historical feature vector of each electric energy meter; Trend module: Based on the periodic statistical vector and periodic law vector, determine the trend value of each statistical feature and each law feature of each electric energy meter, and determine the periodic trend vector of each electric energy meter; Fault module: Determine the real-time feature vector based on the real-time electric energy data, and perform fault diagnosis on all electric energy meters of the monitoring terminal based on the real-time feature vector and the periodic trend vectors of all electric energy meters, and determine the fault electric energy data of each electric energy meter; Interaction module: Display the corresponding fault electric energy data on each electric energy meter of the monitoring terminal and issue an alarm.

[0018] In this embodiment, data such as voltage, power, and electricity consumption of the electric energy meter are collected in real time through Internet of Things devices, and the frequency may be per second or per minute to ensure the capture of instantaneous changes. The data is transmitted to the cloud or a local server through the network. At the same time, historical data (such as daily records for the past few months) including voltage fluctuations, power changes, etc. are extracted from the database to provide a benchmark for subsequent analysis. The collection process needs to ensure time synchronization and data integrity, and the historical data needs to be cleaned to remove missing values.

[0019] In this embodiment, statistical features (such as average voltage, power variance) and law features (such as voltage fluctuation period, electricity consumption trend) are extracted for each electric energy meter cycle by cycle. For example, a certain periodic statistical vector may be [230V, 2kW²], and the law vector is [24h fluctuation, 0.5kWh / day growth]. The two are integrated to form a historical feature vector (such as [230V, 2kW², 24h, 0.5kWh / day]).

[0020] In this embodiment, real-time feature vectors (such as [225V, 1.5kW, 6 times / h]) are extracted from real-time data, and combined with the normal range predicted by the trend vector (such as voltage [219V, 223V]). If the real-time feature exceeds the range (such as 225V), a fault feature vector (such as [225V]) is generated, and further fault power data is output.

[0021] In this embodiment, the fault information is presented on the display device of the monitoring terminal (such as the power meter screen or the management interface), and the user is reminded through sound and light alarms (such as a buzzer or a flashing light). The display content may include the fault type, time, and recommended operations, and the alarm method is adjusted according to the severity (such as high-risk faults triggering remote notifications).

[0022] Beneficial effects of the above technical solution: By analyzing the obtained historical power data, determine the historical feature vectors of each power meter, all statistical features, the trend values of all regular features, and the periodic trend vectors. According to the real-time power data and the periodic trend vectors of all power meters, determine the fault power data of each power meter, display the fault power data of each power meter on the power meter and issue an alarm, which can improve the comprehensiveness of the data, comprehensively capture the dynamic characteristics of the power meter operation, accurately reflect the change law of the characteristics, improve the recognition ability of faults or trend changes, and improve the accuracy, efficiency, and adaptability of power meter fault diagnosis.

[0023] Embodiment 2: The embodiment of the present invention provides an Internet of Things-based remote monitoring and fault diagnosis system for power meters, and the monitoring module includes: Monitoring equipment group unit: Based on the distribution location of the power meters of the monitoring terminal and the monitoring requirements, determine the data monitoring equipment group and the installation location of each data monitoring equipment in the data monitoring equipment group; Real-time power sub-data unit: Based on the installation locations of all data monitoring equipment in the data monitoring equipment group, install the data monitoring equipment group, and based on the installed data monitoring equipment group, collect the real-time power sub-data of each power meter of the monitoring terminal in the real-time monitoring period. The real-time power sub-data includes at least real-time voltage data, real-time power data, and real-time power consumption data; Real-time power data unit: Preprocess the real-time power sub-data of each power meter of the monitoring terminal, and based on the preprocessed real-time power sub-data of all power meters of the monitoring terminal, determine the real-time power data; Historical power sub-data unit: Extract the historical power sub-data of each power meter of the monitoring terminal in multiple historical monitoring periods from the database. Among them, the historical power sub-data includes multiple historical period data, and the historical period data includes periodic voltage data, periodic power data, and periodic power consumption data; Historical Electric Energy Data Unit: Determine historical electric energy data based on the extracted historical electric energy sub-data of all electric energy meters of the monitoring terminal.

[0024] In this embodiment, first analyze the distribution characteristics of the electric energy meters in the monitoring terminal (such as a distribution substation or a factory), such as the number of electric energy meters, spatial locations (such as hierarchical or along the line distribution), and monitoring requirements (such as high-precision voltage monitoring or comprehensive power coverage). Based on this information, design a group of data monitoring devices, which may include devices such as voltage sensors, current sensors, and power collectors, and specify specific installation locations for each device (such as near the main connection terminal of the electric energy meter or key nodes). This design takes into account the monitoring range, data accuracy, and cooperation among devices to ensure comprehensive coverage and minimum interference during data collection. For example, if the electric energy meters are dense in a certain area, the device density may be increased; if the requirement focuses on real-time performance, devices with fast response speeds are preferred. The selection of the installation location also considers the convenience of wiring and the stability of signal transmission.

[0025] In this embodiment, after the location of the monitoring device group unit is determined, physical installation is carried out. For example, the sensors are fixed near the electric energy meters or embedded in the line nodes, and the devices are networked (such as connected to the cloud through the Internet of Things protocol). After the installation is completed, the device group collects electric energy data within the real-time monitoring period at a fixed frequency (such as once per second or once per minute). The real-time electric energy sub-data includes at least three basic indicators: real-time voltage data (reflecting voltage fluctuations), real-time power data (characterizing the load situation), and real-time electricity consumption data (recording the cumulative electricity consumption). The time synchronization and data integrity need to be ensured during the collection process.

[0026] In this embodiment, preprocess the real-time electric energy sub-data, such as denoising (smoothing the data through a filter), filling in missing values (completing them through interpolation), or format standardization (unifying units and time intervals). After preprocessing, integrate the sub-data of all electric energy meters to form real-time electric energy data.

[0027] In this embodiment, retrieve historical data from the system database, covering multiple monitoring periods (such as the past month or year, and each period may be an hour, a day, or a week). The historical electric energy sub-data is a set of records of each electric energy meter in each period, and each group of historical period data includes period voltage data, period power data, and period electricity consumption data.

[0028] In this embodiment, integrate and process the historical electric energy sub-data of all electric energy meters to form unified historical electric energy data.

[0029] Advantages of the above technical solution: Real-time collect the real-time electric energy data of all electric energy meters of the monitoring terminal, obtain the historical electric energy data of all electric energy meters of the monitoring terminal, and provide data basis for determining the historical feature vectors, periodic trend vectors, and real-time feature vectors of each electric energy meter.

[0030] Example 3: The embodiment of the present invention provides an electric energy meter remote monitoring and fault diagnosis system based on the Internet of Things. The analysis module includes: Period statistical vector unit: Extract statistical features from each historical cycle data in the historical electric energy sub-data of each electric energy meter in the historical electric energy data, and determine the period statistical vector of each electric energy meter in each historical monitoring cycle; Voltage-power double broken line graph unit: Based on the period voltage data and period power data in each historical cycle data in the historical electric energy sub-data of each electric energy meter in the historical electric energy data, draw the voltage-power double broken line graph of each electric energy meter in each historical monitoring cycle; Electric energy broken line graph unit: Based on the period electric energy data in each historical cycle data in the historical electric energy sub-data of each electric energy meter in the historical electric energy data, draw the electric energy broken line graph of each electric energy meter in each historical monitoring cycle; Period rule vector unit: Extract rule features from the voltage-power double broken line graph and the electric energy broken line graph of each electric energy meter in each historical monitoring cycle, and determine the period rule vector of each electric energy meter in each historical monitoring cycle; Historical feature vector unit: Determine the historical feature vector of each electric energy meter based on the period statistical vector and the period rule vector of each electric energy meter.

[0031] In this embodiment, the historical electric energy sub-data is analyzed. Taking an electric energy meter as an example, assume there are 30 historical cycles (daily data), and each cycle contains multiple sets of voltage, power, and electric energy values. The unit performs statistical processing on the data of each cycle, such as calculating the average voltage (e.g., daily average value), voltage standard deviation (reflecting fluctuations), maximum and minimum power values (characterizing the load range), etc. These statistical features form a multi-dimensional vector. For example, the statistical vector of a certain cycle may be [average voltage, voltage standard deviation, maximum power, average electric energy]. The processing process may involve outlier removal (such as removing voltage spikes outside the reasonable range) to ensure that the features accurately reflect the cycle characteristics. Finally, each electric energy meter has a corresponding period statistical vector in each cycle, comprehensively describing its numerical distribution characteristics.

[0032] In this embodiment, visual charts are generated using historical cycle data (such as daily voltage and power sequences). The double broken line graph has time as the horizontal axis, and the two broken lines respectively represent the change trends of voltage and power. For example, if the voltage fluctuates from 220V to 230V and the power rises from 5kW to 6kW, the chart will show two curves, revealing the relationship between the two over time. The drawing process may include data smoothing (such as moving average) to highlight the trend, or marking key points (such as voltage mutations). A graph is generated for each cycle of each electric energy meter, reflecting the dynamic relationship between voltage and power, such as the pattern that power increases as voltage rises.

[0033] In this embodiment, a single line graph is plotted with time as the horizontal axis. For example, if the power consumption of a certain electricity meter gradually increases from 1 kWh to 3 kWh, the line graph will show this upward trend. During plotting, data may be preprocessed, such as removing recording errors (such as negative values), or standardizing units (such as unifying to kWh).

[0034] In this embodiment, the patterns in the chart are analyzed to extract regular features. For example, from a voltage-power double line graph, periodic fluctuations (such as a voltage drop and power increase during peak hours every day) or correlations (such as a linear change in power when the voltage is stable) are identified; from the electricity line graph, trends (such as a periodic increase in electricity consumption per week) or outliers (such as a sudden decrease in electricity consumption on a certain day) are extracted. The extraction methods may include time series analysis (such as using Fourier transform to detect periodicity) or pattern recognition (such as marking abnormal inflection points). The result forms a periodic law vector, for example, [voltage fluctuation period, power peak time, electricity growth rate], quantifying the operation law of each period. Each electricity meter obtains a law vector in each period, reflecting its dynamic characteristics.

[0035] In this embodiment, the aforementioned vectors are integrated to form a comprehensive description. For example, the periodic statistical vector of a certain electricity meter may be [average voltage = 220V, voltage standard deviation = 5V, maximum power = 6kW], and the periodic law vector may be [voltage fluctuation period = 24h, power peak time = 18:00, electricity growth rate = 0.5kWh / day]. The two are combined into the historical feature vector of this electricity meter, such as [220V, 5V, 6kW, 24h, 18:00, 0.5kWh / day].[[]]

[0036] Beneficial effects of the above technical solution: By analyzing historical electricity data, determining the periodic statistical vector and periodic law vector of each electricity meter of the monitoring terminal, and determining the historical feature vector of each electricity meter, the dynamic characteristics of the electricity meter operation can be comprehensively captured, the feature expression ability in complex scenarios can be improved, and accurate support can be provided for subsequent trend analysis and fault diagnosis.

[0037] Embodiment 4: The embodiment of the present invention provides an Internet of Things-based remote monitoring and fault diagnosis system for electricity meters. The trend module includes: Trend unit: Analyze the periodic statistical vector and periodic law vector of each electricity meter in all historical monitoring periods, and determine the trend value of each statistical feature in the periodic statistical vector of each electricity meter and the trend value of each law feature in the periodic law vector. Periodic trend vector unit: Based on the trend values of all statistical features in the periodic statistical vector of each electricity meter and the trend values of all law features in the periodic law vector, determine the periodic trend vector of each electricity meter.

[0038] In this embodiment, the periodic trend vector integrates the multi-dimensional vectors of all statistical and regular feature trend values, and describes the dynamic trend of the electricity meter.

[0039] In this embodiment, each electricity meter corresponds to a periodic trend vector.

[0040] Beneficial effects of the above technical solution: Based on the periodic statistical vector and the periodic regular vector, determine the trend values of each statistical feature and each regular feature of each electricity meter, and determine the periodic trend vector of each electricity meter, which can accurately reflect the characteristic change law, improve the recognition ability of complex trends, and provide strong support for fault diagnosis.

[0041] Embodiment 5: The embodiment of the present invention provides a remote monitoring and fault diagnosis system for electricity meters based on the Internet of Things. The trend value unit includes: The first calculation unit: Calculate the trend value of each statistical feature in the periodic statistical vector of each electricity meter based on the periodic statistical vectors of each electricity meter in all historical monitoring periods; ; ; ; ; Wherein, represents the trend value of the a-th statistical feature in the periodic statistical vector of the i-th electricity meter, represents the change trend of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the sign of the change trend of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the trend reliability value of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the first trend reliability threshold, represents the second trend reliability threshold, represents the average feature change value of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, N1 represents the number of extracted historical monitoring periods, represents the feature value of the a-th statistical feature in the periodic statistical vector of the j-th historical monitoring period of the i-th electricity meter, represents the feature value of the a-th statistical feature in the periodic statistical vector of the j+1-th historical monitoring period of the i-th electricity meter, represents the average feature value of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the monitoring time of the j-th historical monitoring period, represents the average monitoring time of all historical monitoring periods, The second calculation unit: calculates the trend value of each regularity feature in the periodic regularity vector of each electricity meter based on the periodic regularity vector of each electricity meter in each historical monitoring period; ; ; ; ; represents the change trend of the b-th regularity feature in the periodic regularity vector of the i-th electricity meter, represents the change trend of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, represents the sign of the change trend of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, represents the trend reliability value of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, represents the average feature change value of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, represents the feature value of the b-th regularity feature in the periodic regularity vector of the j-th historical monitoring period of the i-th electricity meter, represents the feature value of the b-th regularity feature in the periodic regularity vector of the (j + 1)-th historical monitoring period of the i-th electricity meter, represents the average feature value of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter.

[0042] In this embodiment, the monitoring time of the historical monitoring period may be the earliest monitoring time, the latest monitoring time, the intermediate monitoring time, etc. of the historical monitoring period.

[0043] In this embodiment, takes a value of +1 or -1.

[0044] In this embodiment, represents that the trend reliability value of the a-th statistical feature in the periodic statistical vector of all historical monitoring periods of the i-th electricity meter is high, and the change trend is a strong trend, may take a value of 0.6.

[0045] In this embodiment, It indicates that the trend reliability value of the a-th statistical feature in the periodic statistical vector of all historical monitoring periods of the i-th electricity meter is medium, and the change trend is a random trend. The value of

[0046] In this embodiment, It indicates that the trend reliability value of the a-th statistical feature in the periodic statistical vector of all historical monitoring periods of the i-th electricity meter is low, and the change trend is stable.

[0047] Beneficial effects of the above technical solution: By analyzing the periodic statistical vector and the periodic rule vector of each electricity meter in all historical monitoring periods, determining the trend value of each statistical feature in the periodic statistical vector of each electricity meter and the trend value of each rule feature in the periodic rule vector, the change of features over time can be quantified, providing high-quality data support for determining the periodic trend vector of each electricity meter.

[0048] Embodiment 6: The embodiment of the present invention provides an Internet of Things-based remote monitoring and fault diagnosis system for electricity meters. The fault module includes: Real-time statistical vector unit: Extract statistical features from the real-time electricity sub-data of each electricity meter in the real-time electricity data to determine the real-time statistical vector of each electricity meter in the real-time monitoring period; Line graph unit: Draw a real-time first double line graph based on the real-time voltage data and the real-time power data in the real-time electricity sub-data of each electricity meter in the real-time electricity data. At the same time, draw a real-time second line graph based on the real-time electricity quantity data in the real-time electricity sub-data of each electricity meter in the real-time electricity data; Real-time rule vector unit: Extract rule features from the real-time first line graph and the real-time second line graph of each electricity meter to determine the real-time rule vector of each electricity meter in the real-time monitoring period; Real-time feature vector unit: Determine the real-time feature vector based on the real-time statistical vector and the real-time rule vector of each electricity meter in the real-time monitoring period.

[0049] In this embodiment, real-time electricity sub-data is processed and statistically analyzed for the current monitoring period (such as one hour or one day). Taking an electricity meter as an example, assuming that multiple groups of data are collected within the real-time period, the unit calculates statistical indicators, such as the average value of the real-time voltage (reflecting the current voltage level), the standard deviation of the power (characterizing the load fluctuation), and the total amount of electricity (indicating the electricity consumption within the period). These indicators form a real-time statistical vector, such as [average voltage = 225V, power standard deviation = 1.5kW, total electricity = 5kWh]. During processing, noise may be filtered out (such as instantaneous voltage spikes) or invalid values may be excluded (such as negative power) to ensure that the results reflect the true operating state. Each electricity meter generates a vector within the real-time period to quantify its current numerical characteristics.

[0050] In this embodiment, two types of charts are generated. The first double-line chart has time as the horizontal axis (such as data for each minute within one hour), and the two lines respectively show the changes in real-time voltage and power. For example, the voltage rises from 220V to 230V, and the power drops from 3kW to 2kW. The chart reveals the dynamic relationship between the two. The second line chart is a single-line chart that shows the change in real-time electricity, such as the electricity accumulating from 0kWh to 5kWh within the period. During the plotting process, the data may be smoothed (removing jitters) or abnormal points may be marked (such as a sudden increase in power). Each electricity meter generates two charts within the real-time period, visually presenting the voltage-power correlation and the electricity trend, providing a visual basis for pattern extraction.

[0051] In this embodiment, the regular features in the charts are analyzed. Patterns are extracted from the first double-line chart, such as the voltage fluctuation frequency (such as once every 10 minutes) or the correlation between power and voltage (such as power rising when voltage drops); the electricity change trend (such as linear growth) or abnormal points (such as a sudden stop in electricity) are identified from the second line chart. Extraction may detect periodicity through time series analysis or mark abnormalities through inflection point identification. The result forms a real-time regular vector, such as [voltage fluctuation frequency = 6 times / h, power change trend = decreasing, electricity growth rate = 0.1kWh / min], quantifying the operating rules of the current period. Each electricity meter generates a vector to reflect the real-time dynamic characteristics.

[0052] In this embodiment, the statistical and regular features are spliced to form a comprehensive description. The real-time feature vector comprehensively characterizes the current state of the electricity meter and can be compared with historical data for fault detection.

[0053] Beneficial effects of the above technical solution: Determining the real-time feature vector based on real-time electricity data can improve the ability to identify faults or trend changes, providing precise support for rapid fault diagnosis.

[0054] Embodiment 7: The embodiment of the present invention provides an electricity meter remote monitoring and fault diagnosis system based on the Internet of Things. The fault module further includes: Eigenvalue range unit: Based on the historical feature vectors of each electricity meter and the periodic trend vector in the historical monitoring period closest to the real-time monitoring period with the extracted distance, determine the eigenvalue range of each statistical feature and each regular feature of each electricity meter in the real-time monitoring period; Judgment unit: Judge whether each statistical feature and each regular feature in the real-time feature vector of each electricity meter in the real-time monitoring period are within the corresponding eigenvalue range; Fault feature vector unit: For any statistical feature or regular feature that is not within the eigenvalue range, based on all statistical features and all regular features of each electricity meter that are not within the eigenvalue range, determine the fault feature vector of each electricity meter. Otherwise, determine that the fault feature vector of the electricity meter is empty; Fault electricity data unit: Determine the fault electricity data of each electricity meter based on the fault feature vector of each electricity meter.

[0055] In this embodiment, taking the data of the most recent historical period (such as the previous day) as a reference, analyze the historical feature vectors of each electricity meter. Predict the normal range of the real-time period based on this data.

[0056] In this embodiment, compare the real-time feature vector with the eigenvalue range. For example, if the voltage range is [219V, 223V], 225V is out of range; the power standard deviation range is [1.8kW², 2.4kW²], and 1.5kW² is not within the range; the fluctuation frequency range is [5 times / h, 7 times / h], and 6 times / h is normal. The judgment is carried out item by item to check whether each feature falls within the expected range.

[0057] In this embodiment, if a certain electricity meter has features that exceed the range, the unit extracts these abnormal features to form a fault feature vector. For example, if the real-time feature vector is [225V, 1.5kW, 6 times / h], the voltage range [219V, 223V] and the power standard deviation range [1.8kW², 2.4kW²] are exceeded, and the fault feature vector is [225V, 1.5kW] (only including abnormal items). If all features are within the range, such as [220V, 2kW, 6 times / h], the fault feature vector is empty (indicating no fault).

[0058] In this embodiment, generate a fault description according to the fault feature vector, and record in detail the abnormal features and their deviations. If the fault feature vector is empty, the data is empty or marked as "no fault". The generation process includes feature mapping (such as voltage abnormality corresponding to line problems) or data formatting (for system storage or display), providing a basis for fault location and repair.

[0059] Beneficial effects of the above technical solution: Based on the real-time feature vectors and the periodic trend vectors of all electricity meters, fault diagnosis is performed on all electricity meters of the monitoring terminal to determine the faulty electricity data of each electricity meter, which can accurately identify faults, improve the accuracy and efficiency of fault diagnosis in complex scenarios, and provide strong support for the rapid location and repair of electricity meter faults.

[0060] Embodiment 8: The embodiment of the present invention provides an Internet of Things-based remote monitoring and fault diagnosis system for electricity meters. The eigenvalue range unit includes: ; ; ; ; wherein, represents the eigenvalue range of the a-th statistical feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter, represents the predicted eigenvalue of the a-th statistical feature in the periodic statistical vector of the real-time monitoring period of the i-th electricity meter, represents the eigenvalue standard deviation of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the eigenvalue of the a-th statistical feature in the periodic statistical vector of the N1-th historical monitoring period of the i-th electricity meter, represents the monitoring time of the real-time monitoring period, represents the monitoring time of the N1-th historical monitoring period, represents the eigenvalue range of the b-th regular feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter, represents the predicted eigenvalue of the b-th regular feature in the periodic regular vector of the real-time monitoring period of the i-th electricity meter, represents the eigenvalue of the b-th regular feature in the periodic regular vector of the N1-th historical monitoring period of the i-th electricity meter, represents the eigenvalue standard deviation of the b-th regular feature in the periodic regular vectors of all historical monitoring periods of the i-th electricity meter.

[0061] In this embodiment, represents the lower limit of the eigenvalue range of the a-th statistical feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter.

[0062] In this embodiment, represents the upper limit of the eigenvalue range of the a-th statistical feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter.

[0063] In this embodiment, Represents the lower limit of the eigenvalue range of the b-th regular feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter.

[0064] In this embodiment, Represents the upper limit of the eigenvalue range of the b-th regular feature in the real-time feature vector of the real-time monitoring period of the i-th electricity meter.

[0065] Beneficial effects of the above technical solution: Determining the eigenvalue range of each statistical feature and each regular feature of each electricity meter in the real-time monitoring period can improve data comprehensiveness, comprehensively capture the dynamic characteristics of the electricity meter operation, accurately reflect the feature change law, improve the recognition ability of faults or trend changes, and improve the accuracy, efficiency and adaptability of electricity meter fault diagnosis.

[0066] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0067] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0068] 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 them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things, characterized in that, including: Monitoring module: Collecting in real time the real-time power data of all the watt-hour meters of the monitoring terminal, and obtaining the historical power data of all the watt-hour meters of the monitoring terminal; Analysis module: Analyzing the historical power data, determining the periodic statistical vector and the periodic pattern vector of each watt-hour meter of the monitoring terminal, and determining the historical feature vector of each watt-hour meter; Trend module: Based on the periodic statistical vector and the periodic pattern vector, determining the trend value of each statistical feature and each pattern feature of each watt-hour meter, and determining the periodic trend vector of each watt-hour meter; Fault module: Determining the real-time feature vector based on the real-time power data, and performing fault diagnosis on all the watt-hour meters of the monitoring terminal based on the real-time feature vector and the periodic trend vectors of all the watt-hour meters, and determining the fault power data of each watt-hour meter; Interaction module: Displaying the corresponding fault power data on each watt-hour meter of the monitoring terminal and sending out an alarm.

2. The remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things according to claim 1, characterized in that, Monitoring module, including: Monitoring device group unit: Determining the data monitoring device group and the installation location of each data monitoring device in the data monitoring device group based on the distribution location of the watt-hour meters of the monitoring terminal and the monitoring requirements; Real-time power sub-data unit: Installing the data monitoring device group based on the installation locations of all the data monitoring devices in the data monitoring device group, and collecting in real time the real-time power sub-data of each watt-hour meter of the monitoring terminal during the real-time monitoring period based on the installed data monitoring device group. The real-time power sub-data includes at least real-time voltage data, real-time power data, and real-time power consumption data; Real-time power data unit: Preprocessing the real-time power sub-data of each watt-hour meter of the monitoring terminal, and determining the real-time power data based on the preprocessed real-time power sub-data of all the watt-hour meters of the monitoring terminal; Historical power sub-data unit: Extracting the historical power sub-data of each watt-hour meter of the monitoring terminal in multiple historical monitoring periods from the database. The historical power sub-data includes multiple historical period data, and the historical period data includes periodic voltage data, periodic power data, and periodic power consumption data; Historical power data unit: Determining the historical power data based on the extracted historical power sub-data of all the watt-hour meters of the monitoring terminal.

3. The remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things according to claim 2, characterized in that, Analysis module, including: Periodic statistical vector unit: Extracting statistical features from each historical period data in the historical power sub-data of each watt-hour meter in the historical power data, and determining the periodic statistical vector of each watt-hour meter in each historical monitoring period; Voltage-power double broken line graph unit: Drawing the voltage-power double broken line graph of each watt-hour meter in each historical monitoring period based on the periodic voltage data and the periodic power data in each historical period data in the historical power sub-data of each watt-hour meter in the historical power data; Power consumption broken line graph unit: Drawing the power consumption broken line graph of each watt-hour meter in each historical monitoring period based on the periodic power consumption data in each historical period data in the historical power sub-data of each watt-hour meter in the historical power data; Periodic pattern vector unit: Extracting pattern features from the voltage-power double broken line graph and the power consumption broken line graph of each watt-hour meter in each historical monitoring period, and determining the periodic pattern vector of each watt-hour meter in each historical monitoring period; Historical feature vector unit: Determine the historical feature vector of each electricity meter based on the periodic statistical vector and periodic pattern vector of each electricity meter.

4. The remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things according to claim 3, characterized in that, Trend module, including: Trend unit: Analyze the periodic statistical vector and periodic pattern vector of each electricity meter in all historical monitoring periods, and determine the trend value of each statistical feature in the periodic statistical vector of each electricity meter and the trend value of each pattern feature in the periodic pattern vector; Periodic trend vector unit: Determine the periodic trend vector of each electricity meter based on the trend values of all statistical features in the periodic statistical vector of each electricity meter and the trend values of all pattern features in the periodic pattern vector.

5. The remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things according to claim 3, characterized in that, Trend value unit, including: First calculation unit: Calculate the trend value of each statistical feature in the periodic statistical vector of each electricity meter based on the periodic statistical vector of each electricity meter in all historical monitoring periods; ； ； ； ； Among them, represents the trend value of the a-th statistical feature in the periodic statistical vector of the i-th electricity meter, represents the change trend of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the sign of the change trend of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the trend reliability value of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the first trend reliability threshold, represents the second trend reliability threshold, represents the average feature change value of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, and N1 represents the number of extracted historical monitoring periods, represents the feature value of the a-th statistical feature in the periodic statistical vector of the j-th historical monitoring period of the i-th electricity meter, represents the feature value of the a-th statistical feature in the periodic statistical vector of the (j + 1)-th historical monitoring period of the i-th electricity meter, represents the average feature value of the a-th statistical feature in the periodic statistical vectors of all historical monitoring periods of the i-th electricity meter, represents the monitoring time of the j-th historical monitoring period, represents the average monitoring time of all historical monitoring periods, Second calculation unit: Calculate the trend value of each pattern feature in the periodic pattern vector of each electricity meter based on the periodic pattern vector of each electricity meter in each historical monitoring period; ； ； ； ； Represents the change trend of the b-th regularity feature in the periodic regularity vector of the i-th electricity meter, Represents the change trend of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, Represents the sign of the change trend of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, Represents the trend reliability value of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, Represents the average feature change value of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter, Represents the feature value of the b-th regularity feature in the periodic regularity vector of the j-th historical monitoring period of the i-th electricity meter, Represents the feature value of the b-th regularity feature in the periodic regularity vector of the (j + 1)-th historical monitoring period of the i-th electricity meter, Represents the average feature value of the b-th regularity feature in the periodic regularity vectors of all historical monitoring periods of the i-th electricity meter.

6. The remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things according to claim 3, characterized in that, Fault module, including: Real-time statistical vector unit: Extract statistical features from the real-time electricity sub-data of each electricity meter in the real-time electricity data, and determine the real-time statistical vector of each electricity meter in the real-time monitoring period; Line chart unit: Draw a real-time first double line chart based on the real-time voltage data and real-time power data in the real-time electricity sub-data of each electricity meter in the real-time electricity data. At the same time, draw a real-time second line chart based on the real-time electricity data in the real-time electricity sub-data of each electricity meter; Real-time pattern vector unit: Extract pattern features from the real-time first line chart and real-time second line chart of each electricity meter, and determine the real-time pattern vector of each electricity meter in the real-time monitoring period; Real-time feature vector unit: Determine the real-time feature vector based on the real-time statistical vector and real-time pattern vector of each electricity meter in the real-time monitoring period.

7. The remote monitoring and fault diagnosis system for electric energy meters based on the Internet of Things according to claim 3, characterized in that, The fault module also includes: Eigenvalue range unit: Determine the eigenvalue range of each statistical feature and each pattern feature of each electricity meter in the real-time monitoring period based on the historical feature vector and periodic trend vector of each electricity meter in the historical monitoring period closest to the real-time monitoring period; Judgment unit: Judge whether each statistical feature and each pattern feature in the real-time feature vector of each electricity meter in the real-time monitoring period are within the corresponding eigenvalue range; Fault feature vector unit: For an electricity meter with any statistical feature or pattern feature not within the eigenvalue range, determine the fault feature vector of each electricity meter based on all statistical features and all pattern features of the electricity meter not within the eigenvalue range. Otherwise, determine that the fault feature vector of the electricity meter is empty; Fault electricity data unit: Determine the fault electricity data of each electricity meter based on the fault feature vector of each electricity meter.

8. The remote monitoring and fault diagnosis system for watt-hour meters based on the Internet of Things according to claim 7, wherein Eigenvalue range unit, including: ； ； ； ； Among them, represents the eigenvalue range of the ath statistical feature in the real-time feature vector of the real-time monitoring period of the ith electricity meter, represents the predicted eigenvalue of the ath statistical feature in the periodic statistical vector of the real-time monitoring period of the ith electricity meter, represents the standard deviation of the eigenvalues of the ath statistical feature in the periodic statistical vectors of all historical monitoring periods of the ith electricity meter, represents the eigenvalue of the ath statistical feature in the periodic statistical vector of the N1th historical monitoring period of the ith electricity meter, represents the monitoring time of the real-time monitoring period, represents the monitoring time of the N1th historical monitoring period, represents the eigenvalue range of the bth regular feature in the real-time feature vector of the real-time monitoring period of the ith electricity meter, represents the predicted eigenvalue of the bth regular feature in the periodic regular vector of the real-time monitoring period of the ith electricity meter, represents the eigenvalue of the bth regular feature in the periodic regular vector of the N1th historical monitoring period of the ith electricity meter, represents the standard deviation of the eigenvalues of the bth regular feature in the periodic regular vectors of all historical monitoring periods of the ith electricity meter.

Images