Online Fault Monitoring and Analysis of Electricity Meters Based on Multi-Source Data Fusion

Abstract

This invention relates to the field of electricity meter fault monitoring technology, specifically to an electricity meter for online fault monitoring and analysis based on multi-source data fusion. The electricity meter includes: a data acquisition module that acquires temperature and contact resistance sequences of the trip switch contacts of the target electricity meter and other electricity meters, and sets monitoring windows; an independent resistance anomaly analysis module that obtains the resistance anomaly value of the current monitoring window; a relative anomaly analysis module that records the monitoring windows corresponding to other electricity meters as comparison windows; obtains the comparability evaluation value and the reliability of the comparison results for each comparison window; obtains relative anomaly indicators using the resistance anomaly values ​​of the comparison windows and the current monitoring window; and an early warning monitoring module that acquires each reference window of the current monitoring window; obtains the relative anomaly indicator density value of the current monitoring window using each reference window; and then obtains early warning indicators and issues early warnings based on these indicators. This application can effectively monitor electricity meter faults.

Description

Technical Field

[0001] This invention relates to the field of electricity meter fault monitoring technology, specifically to an electricity meter for online fault monitoring and analysis based on multi-source data fusion. Background Technology

[0002] As power systems continue to develop towards intelligence and precision, electricity meters, as crucial metering and monitoring devices at the end of the distribution network, directly impact the accuracy of electricity measurement and the safety of power supply and consumption. Electricity meters typically possess trip protection functions to control circuit continuity. The contact state and closing stability of their trip switch contacts significantly affect power supply continuity and operational safety. In actual operation, electricity meters operate in complex electrical environments, susceptible to high loads and frequent power interruptions, potentially leading to oxidation and carbon buildup on the trip switch contacts. This can affect contact performance, causing decreased power supply stability and even safety hazards such as abnormal overheating. Therefore, monitoring the contact state and performance stability of critical components like trip switches is essential for ensuring reliable electricity meter operation. Continuous monitoring and analysis of electrical parameters and related status information generated during electricity meter operation helps reflect changes in the working characteristics and behaviors of internal components, providing data support for assessing meter operating status and judging performance trends of key components, thereby improving the accuracy and timeliness of electricity meter operation management.

[0003] In the performance monitoring of electricity meter trip switches, existing monitoring methods typically analyze the trip switch status based on a single operational data point, such as judging its contact condition based on the magnitude of contact resistance. However, in actual operation, trip switches are easily affected by factors such as load changes, frequent power on / off cycles, and heat generation, and related electrical parameters may fluctuate dynamically. A single data point cannot fully reflect the true working state of the trip switch, thus affecting the stability and reliability of the performance evaluation results. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide an online fault monitoring and analysis energy meter based on multi-source data fusion. The specific technical solution adopted is as follows:

[0005] One embodiment of the present invention provides an online fault monitoring and analysis energy meter based on multi-source data fusion, the energy meter comprising:

[0006] The data acquisition module is used to collect the temperature sequence and contact resistance sequence of the trip switch contacts of the target energy meter and other energy meters that are identical to the target energy meter; and to set up a monitoring window to divide each temperature sequence and contact resistance sequence.

[0007] The independent resistance anomaly analysis module is used to obtain the resistance anomaly value of the current monitoring window based on each contact resistance in the current monitoring window of the target energy meter's contact resistance sequence.

[0008] The relative anomaly analysis module is used to designate the monitoring windows corresponding to other energy meters that are identical to the target energy meter as comparison windows; obtain the comparability evaluation value of the comparison window based on the temperature difference between the current monitoring window and a comparison window; obtain the reliability of the comparison result of the comparison window based on the comparability evaluation value of a comparison window and the commissioning time of the energy meter corresponding to the comparison window; and obtain the relative anomaly index of the current monitoring window using the resistance anomaly value of the current monitoring window, the resistance anomaly values ​​of each comparison window, and the reliability of the comparison result.

[0009] The early warning monitoring module is used to obtain each reference window of the current monitoring window from the monitoring windows before the current monitoring window of the target energy meter using relative anomaly indicators; to obtain the relative anomaly indicator density value of the current monitoring window using the relative anomaly indicators of the current monitoring window, the relative anomaly indicators of each reference window, and the time distance between each reference window and the current monitoring window; to fuse the relative anomaly indicator density value and the relative anomaly indicators of the current monitoring window to obtain an early warning indicator; and to issue an early warning based on the early warning indicator.

[0010] Preferably, the abnormal resistance value of the current monitoring window is obtained based on each contact resistance within the current monitoring window in the contact resistance sequence of the target energy meter, including:

[0011] The coefficient of variation of each contact resistance in the current monitoring window of the contact resistance sequence of the target energy meter is mapped using an exponential function with the natural constant as the base to obtain a first mapped value. The resistance anomaly value of the current monitoring window is obtained by multiplying the first mapped value by the ratio of the mean value of each contact resistance in the current monitoring window of the contact resistance sequence of the target energy meter to the reference value of the contact resistance of the target energy meter.

[0012] Preferably, obtaining the comparability assessment value of the comparison window based on the temperature difference between the current monitoring window and a comparison window includes:

[0013] A first temperature similarity is obtained by negatively correlating the normalized value of the dynamic time-normalized distance between the temperature sequence within the current monitoring window and the temperature sequence within a comparison window using an exponential function with a base of the natural constant. A second temperature similarity is obtained by negatively correlating the normalized value of the absolute value of the difference between the mean temperature within the current monitoring window and the mean temperature within the comparison window using an exponential function with a base of the natural constant. The first and second temperature similarities are then multiplied to obtain the comparability assessment value of the comparison window.

[0014] Preferably, the reliability of the comparison results of a comparison window is obtained based on the comparability evaluation value of a comparison window and the commissioning time of the corresponding electricity meter, including:

[0015] The reliability of the comparison results of a comparison window is obtained by negatively mapping the time distance between the commissioning time of the electricity meter corresponding to a comparison window and the time corresponding to the comparison window using an exponential function with the natural constant as the base, and multiplying it by the comparability evaluation value of the comparison window.

[0016] Preferably, the relative anomaly index of the current monitoring window is obtained by utilizing the resistance anomaly value of the current monitoring window, the resistance anomaly values ​​of each comparison window, and the reliability of the comparison results, including:

[0017] The resistance anomaly value of the current monitoring window is compared with the resistance anomaly value of a comparison window to obtain the resistance anomaly value ratio of the comparison window; the reliability of the comparison results of each comparison window is used as the weight of the resistance anomaly value ratio of each comparison window to obtain the weighted average of the resistance anomaly value ratios of the comparison windows to obtain the relative anomaly index of the current monitoring window.

[0018] Preferably, the reference windows for the current monitoring window are obtained from the monitoring windows preceding the current monitoring window of the target energy meter using relative anomaly indicators, including:

[0019] Obtain a preset number of different monitoring windows from the monitoring windows before the current monitoring window of the target energy meter that have the closest relative abnormality index to the current monitoring window as reference windows for the current monitoring window.

[0020] Preferably, the relative anomaly density value of the current monitoring window is obtained by utilizing the relative anomaly indices of the current monitoring window, the relative anomaly indices of each reference window, and the time distance between each reference window and the current monitoring window, including:

[0021] Obtain the normalized value of the absolute value of the difference between the relative anomaly index of the current monitoring window and a reference window, and denot it as the relative anomaly difference corresponding to the reference window; multiply the relative anomaly difference corresponding to the reference window by the normalized value of the time distance between the reference window and the current monitoring window to obtain the relative anomaly difference of fusion time corresponding to the reference window; use an exponential function with the natural constant as the base to perform a negative correlation mapping on the mean of the relative anomaly differences of fusion time corresponding to each reference window to obtain the relative anomaly index density value of the current monitoring window.

[0022] Preferably, the early warning indicators are obtained by fusing the relative anomaly density values ​​and relative anomaly indicators of the current monitoring window, including:

[0023] A second mapping value is obtained by mapping the relative anomaly index of the current monitoring window to an exponential function with the natural constant as the base; the warning index of the current monitoring window is obtained by mapping the product of the relative anomaly index density value of the current monitoring window and the second mapping value to a sigmoid function.

[0024] The embodiments of the present invention have at least the following beneficial effects: This application collects the temperature sequence and contact resistance sequence of the trip switch contacts of the target energy meter and other energy meters identical to the target energy meter; and sets a monitoring window to divide each temperature sequence and contact resistance sequence; further, based on each contact resistance in the current monitoring window in the contact resistance sequence of the target energy meter, it obtains the resistance anomaly value of the current monitoring window to reflect the instantaneous change and stability of the contact performance; then, the monitoring windows corresponding to other energy meters identical to the target energy meter are recorded as comparison windows, and then the relative anomaly index of the current monitoring window is obtained by using the resistance anomaly value of the current monitoring window, the resistance anomaly values ​​of each comparison window, and the reliability of the comparison results, taking into account temperature and usage time. The application improves the objectivity and robustness of anomaly assessment by considering differences in operating conditions. Finally, it uses relative anomaly indicators to obtain reference windows from previous monitoring windows of the target energy meter, calculates the relative anomaly index density value of the current monitoring window based on these reference windows, and then fuses the relative anomaly index density value and the relative anomaly index to obtain an early warning index. This comprehensively characterizes the persistence and concentration of anomalies over time. Based on this early warning index, an early warning is issued, effectively distinguishing between short-term disturbances and actual contact degradation. Therefore, this application enables real-time monitoring and early warning of tripped switch contact quality, providing a scientific basis for energy meter operation and maintenance decisions and condition-based maintenance, and improving the safety and reliability of the power supply system. Attached Figure Description

[0025] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a module framework diagram of an energy meter for online fault monitoring and analysis based on multi-source data fusion, provided as an embodiment of the present invention. Detailed Implementation

[0027] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an online fault monitoring and analysis energy meter based on multi-source data fusion proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0029] The following description, in conjunction with the accompanying drawings, details a specific scheme for online fault monitoring and analysis of an energy meter based on multi-source data fusion provided by this invention.

[0030] Example:

[0031] The main application scenario of this invention is as follows: This application mainly analyzes the target energy meter by combining the operating data of multiple identical energy meters, thereby monitoring the fault status of the target energy meter.

[0032] Please see Figure 1 The diagram illustrates a module block diagram of an online fault monitoring and analysis energy meter based on multi-source data fusion, according to an embodiment of the present invention. The energy meter includes the following modules:

[0033] The data acquisition module is used to collect the temperature sequence and contact resistance sequence of the trip switch contacts of the target energy meter and other energy meters that are the same as the target energy meter; and to set up a monitoring window to divide each temperature sequence and contact resistance sequence.

[0034] This application primarily aims to monitor and provide early warning of abnormal states in the tripping switches of electricity meters. The electricity meter to be analyzed is designated as the target meter. Operational data from the target meter and other meters of the same model are collected. The collected data includes the commissioning time (initial commissioning time) of both the target and target meters, as well as the temperature and contact resistance sequences of the tripping switch contacts during operation. The time of each data point is marked to ensure the integrity of the timing information. During data collection, the data is uploaded to the terminal system in real time using the electricity meter's built-in data transmission module, ensuring centralized management and analysis of data from all meters on a unified platform. This data is used for subsequent real-time monitoring and accurate early warning of the tripping switch contact quality.

[0035] When the contacts of a trip switch experience poor contact or aging and deterioration, their contact resistance typically increases significantly compared to the initial state of the equipment. Simultaneously, influenced by factors such as contact fretting, thermal expansion, and uneven surface conditions, their contact stability also decreases significantly, exhibiting a characteristic of increased fluctuation over a short period. Based on these characteristics, this solution first pre-defines a unified monitoring window (the monitoring window is a time window, sliding in steps of the window length on the corresponding sequence to achieve real-time monitoring) to ensure the temporal consistency of data analysis. The preferred length of the monitoring window in this application is 1 second. This allows for the division of temperature and contact resistance sequences using the monitoring window, obtaining the data values ​​of the monitoring window on each temperature and contact resistance sequence.

[0036] The independent resistance anomaly analysis module is used to obtain the resistance anomaly value of the current monitoring window based on each contact resistance in the current monitoring window of the target energy meter's contact resistance sequence.

[0037] During long-term operation, the contacts of trip switches in electricity meters are susceptible to damage from factors such as electric arcing, current heating, and mechanical wear under conditions of high load and frequent power on / off cycles. This can lead to oxidation, carbon buildup, or microscopic ablation on the contact surface, weakening the effective contact area between contacts, reducing conductivity reliability, and ultimately resulting in decreased contact performance and reduced power supply stability. Therefore, it is necessary to continuously monitor and quantitatively analyze the external manifestations of poor contact in trip switches and their potential performance degradation to enable timely identification and early warning of abnormal conditions.

[0038] Therefore, the current monitoring window of the target energy meter is recorded as the current monitoring window, and the resistance anomaly value of the current monitoring window is obtained according to each contact resistance in the contact resistance sequence of the target energy meter within the current monitoring window.

[0039] Specifically, a first mapping value is obtained by mapping the coefficient of variation of each contact resistance in the current monitoring window of the contact resistance sequence of the target energy meter to an exponential function with the natural constant as the base; the resistance anomaly value of the current monitoring window is obtained by multiplying the first mapping value by the ratio of the mean value of each contact resistance in the current monitoring window of the contact resistance sequence of the target energy meter to the reference value of the contact resistance of the target energy meter.

[0040] The specific calculation model for abnormal resistance values ​​is as follows:

[0041] ,

[0042] Where Z represents the abnormal resistance value of the target energy meter in the current monitoring window; This represents the average value of the contact resistance collected within the current monitoring window; This indicates the reference value of the contact resistance of the target energy meter under normal temperature conditions during factory testing. This ratio characterizes the degree of deviation of the current contact state from the initial healthy state; the larger the ratio, the more significant the degradation of contact performance. e is the natural constant. This represents the coefficient of variation of each contact resistance within the current monitoring window, used to characterize the contact stability level. The larger the value, the more unstable the contact state. By introducing an exponential mapping method to obtain the first mapping value, the contact resistance fluctuation characteristics are nonlinearly amplified. This avoids feature failure due to the coefficient of variation approaching zero, and enhances the sensitivity to unstable contact states.

[0043] Similarly, the abnormal resistance values ​​of other monitoring windows of the target energy meter and other energy meters that are the same as the target energy meter are the same as the abnormal resistance values ​​of the current monitoring window.

[0044] The relative anomaly analysis module is used to designate the monitoring windows corresponding to other energy meters that are identical to the target energy meter as comparison windows; obtain the comparability evaluation value of the comparison window based on the temperature difference between the current monitoring window and a comparison window; obtain the reliability of the comparison result of the comparison window based on the comparability evaluation value of a comparison window and the commissioning time of the energy meter corresponding to the comparison window; and obtain the relative anomaly index of the current monitoring window using the resistance anomaly value of the current monitoring window, the resistance anomaly values ​​of each comparison window, and the reliability of the comparison result.

[0045] The above method allows for the acquisition of basic anomaly indicators (resistance anomalies) characterizing the contact state of trip switches within any electricity meter and any monitoring window. However, during actual operation of the electricity meter, the physical properties of the contact material change with temperature due to the thermal effect of the current carried by the contacts and the influence of external ambient temperature variations. This results in different characterizations of the contact resistance and its stability under different temperature conditions. Furthermore, as the electricity meter's usage time increases, its overall performance may slowly drift. To reduce the impact of the shift caused by the evolution of a single meter over time, this scheme compares the current monitoring window of the target electricity meter under analysis with the monitoring windows of other electricity meters of the same model, thereby improving the objectivity and robustness of the anomaly assessment results.

[0046] Therefore, the monitoring windows corresponding to other electricity meters that are identical to the target electricity meter are recorded as comparison windows. The current monitoring window has multiple comparison windows.

[0047] When conducting comparative analysis, in order to ensure reasonable physical comparability between the current monitoring window and the comparison window, the comparability assessment value between the current monitoring window and any comparison window is first used to evaluate the abnormal resistance value of the trip switch contact between the current monitoring window and any comparison window.

[0048] Therefore, the comparability assessment value of the comparison window is obtained based on the temperature difference between the current monitoring window and a comparison window. Specifically, a first temperature similarity is obtained by negatively correlating the normalized value of the dynamic time-normalized distance between the temperature sequences in the current monitoring window and the temperature sequences in the comparison window using an exponential function with a base of the natural constant; a second temperature similarity is obtained by negatively correlating the normalized value of the absolute value of the difference between the mean temperature in the current monitoring window and the mean temperature in the comparison window using an exponential function with a base of the natural constant; and the comparability assessment value of the comparison window is obtained by multiplying the first temperature similarity and the second temperature similarity.

[0049] The specific calculation model for the comparability assessment value is as follows:

[0050] ,

[0051] In the formula, This represents the comparability assessment value between the current monitoring window a and a comparison window b, which is also the comparability assessment value of the comparison window; norm() is a linear normalization function used to eliminate the influence of dimensional differences; , These represent the sequences of temperatures at the trip switch contacts collected in the current monitoring window a and the sequences of temperatures at the trip switch contacts collected in the comparison window b, respectively. This represents the dynamic time-warped distance (DTW) between the two temperature sequences within the current monitoring window a and the comparison window b (obtained using the DTW algorithm). It characterizes the similarity of temperature changes; the smaller the DTW distance, the more consistent the temperature change patterns within the two windows, and the stronger the comparative significance of the corresponding contact resistance anomalies. The first temperature similarity; , These are the mean values ​​of the temperature sequences corresponding to the current monitoring window a and the comparison window b, respectively. This indicates the difference in overall contact temperature levels between the current monitoring window a and the comparison window b. The smaller the difference, the closer the operating thermal conditions are, and the higher the comparability. This represents the second temperature similarity.

[0052] The above method allows us to obtain a comparability assessment value between the current monitoring window and any comparison window. In actual comparisons, considering that the probability of a tripped switch causing contact abnormalities is relatively low in the initial stage of use, and its operating state is closer to a healthy baseline state, the comparison results are more reliable when it is used as a comparison window. Therefore, when analyzing the current monitoring window, a time decay factor is introduced to weighted model the reliability of the corresponding comparison window.

[0053] Therefore, the reliability of the comparison results of a comparison window is obtained based on the comparability evaluation value of a comparison window and the commissioning time of the electricity meter corresponding to the comparison window.

[0054] Specifically, the reliability of the comparison results of a comparison window is obtained by negatively mapping the time distance between the commissioning time of the electricity meter corresponding to a comparison window and the time corresponding to the comparison window using an exponential function with the natural constant as the base, and multiplying it by the comparability evaluation value of the comparison window.

[0055] The specific model for calculating the credibility of the comparison results in the comparison window is as follows:

[0056] ,

[0057] In the formula, This represents the evaluation value of the reliability of the comparison result of comparison window b to the current monitoring window a, which is also the reliability of the comparison result of comparison window b. This represents the comparability assessment value of comparison window b. The larger the value, the closer the operating conditions of comparison window b are to the current monitoring window a, and the stronger its reference significance. This represents the time distance between the time corresponding to comparison window b and the time when the electricity meter corresponding to comparison window b was put into operation (the time when the electricity meter started to be used). The time corresponding to comparison window b is the middle time of the window. The smaller the value, the closer the comparison window is to the initial stage of device operation, and the higher its reliability as a health reference.

[0058] Through the above processing, and taking into account the similarity of operating conditions and time reliability, the relative anomaly index of the contact quality of the trip switch is constructed by comparing the data of the current monitoring window of the target energy meter with the monitoring window data of other energy meters as a reference.

[0059] The relative abnormality index of the current monitoring window is obtained by using the resistance abnormality value of the current monitoring window, the resistance abnormality value of each comparison window, and the reliability of the comparison results.

[0060] Specifically, the resistance anomaly value of the current monitoring window is compared with the resistance anomaly value of a comparison window to obtain the resistance anomaly value ratio corresponding to the comparison window; the reliability of the comparison results of each comparison window is used as the weight of the resistance anomaly value ratio corresponding to each comparison window to obtain the weighted average of the resistance anomaly value ratios corresponding to the comparison windows to obtain the relative anomaly index of the current monitoring window.

[0061] The specific calculation model for relative anomaly indicators is as follows:

[0062] ,

[0063] In the formula, This represents the relative abnormality index of the trip switch contact quality within the current monitoring window 'a' of the target energy meter; 'm' represents the number of comparison windows participating in the comparison. , These represent the abnormal resistance values ​​of the contact resistance corresponding to the current monitoring window a and the comparison window b, respectively. This indicates the reliability of the comparison result of comparison window b to the current monitoring window a, which is also the reliability of the comparison result of comparison window b. It is used as the weight of the resistance anomaly value ratio corresponding to the comparison window. The ratio of the resistance anomaly value corresponding to the comparison window is used to characterize the degree of abnormal deviation of the current monitoring window relative to the comparison window. The larger the ratio and the higher the confidence evaluation value of the corresponding comparison window, the more accurate and reliable the judgment result of the current trip switch contact quality anomaly.

[0064] The early warning monitoring module is used to obtain each reference window of the current monitoring window from the monitoring windows before the current monitoring window of the target energy meter using relative anomaly indicators; to obtain the relative anomaly indicator density value of the current monitoring window using the relative anomaly indicators of the current monitoring window, the relative anomaly indicators of each reference window, and the time distance between each reference window and the current monitoring window; to fuse the relative anomaly indicator density value and the relative anomaly indicators of the current monitoring window to obtain an early warning indicator; and to issue an early warning based on the early warning indicator.

[0065] The above analysis yields the relative anomaly index of the trip switch contact quality of the target energy meter within the current monitoring window 'a'. During actual operation of the energy meter, transient abnormal fluctuations may occur within local timeframes due to load changes, environmental disturbances, or momentary switching behavior. However, such transient anomalies do not necessarily correspond to actual contact performance degradation or fault conditions. To avoid misjudgment due to a single momentary anomaly, this application, based on the relative anomaly index obtained within the current monitoring window, further analyzes the persistence and concentration of the anomaly by considering its historical distribution characteristics.

[0066] Therefore, by using relative anomaly indicators, reference windows for the current monitoring window are obtained from the monitoring windows preceding the current monitoring window of the target energy meter.

[0067] Specifically, a predetermined number of different monitoring windows whose relative anomaly indicators are closest to the current monitoring window are obtained from the monitoring windows preceding the current monitoring window of the target energy meter, serving as reference windows for the current monitoring window. This application selects n historical monitoring windows as reference windows, where n is preferably 100. This method ensures that comparative analysis is performed only within historical states with similar degrees of anomaly, thereby avoiding interference from health status data or extreme anomaly data with the evaluation results. Furthermore, the difference relationship and temporal distribution characteristics between each anomaly indicator in the reference windows and the current anomaly indicator are calculated, thus obtaining the density estimate of the anomaly indicator in the current monitoring window.

[0068] Therefore, the relative abnormality index density value of the current monitoring window is obtained by using the relative abnormality index of the current monitoring window, the relative abnormality index of each reference window, and the time distance between each reference window and the current monitoring window.

[0069] Specifically, the normalized value of the absolute value of the difference between the relative anomaly index of the current monitoring window and a reference window is obtained and denoted as the relative anomaly difference corresponding to the reference window; the relative anomaly difference corresponding to the reference window is multiplied by the normalized value of the time distance between the reference window and the current monitoring window to obtain the relative anomaly difference of fusion time corresponding to the reference window; the mean of the relative anomaly differences of fusion time corresponding to each reference window is negatively correlated using an exponential function with the natural constant as the base to obtain the relative anomaly index density value of the current monitoring window.

[0070] The specific calculation model for the relative anomaly index density value is as follows:

[0071] ,

[0072] In the formula, represents the density value of abnormal index of the contact quality of the trip switch within the current monitoring window a of the target energy meter; exp[ ] represents the exponential function with the natural constant as the base, which is used here for the negative correlation mapping of the results; , These represent the relative abnormality indicators of the trip switch contact quality in the current monitoring window a and the h-th reference window in its selected historical reference window, respectively. This represents the time distance between the current monitoring window and the h-th reference window. When calculating the time distance, the midpoint of the window is used. `norm` represents the normalization operation. The above density estimation method considers both the differences in anomaly indicators and the time interval. When the differences in anomaly indicators are small and the time distance is short, their corresponding terms contribute more to the density estimation result, thus reflecting that the current anomaly has the characteristics of continuous occurrence or concentrated distribution in the time dimension. Conversely, when the anomaly only occurs in isolation or differs greatly from historical states, its density estimate is relatively low.

[0073] Through the above processing, the relative anomaly index density value of the trip switch contact quality within the current monitoring window a of the target energy meter can be obtained. Furthermore, this application integrates the relative anomaly index density value with the relative anomaly index within the current monitoring window to construct an early warning index for the trip switch contact quality.

[0074] Specifically, a second mapping value is obtained by mapping the relative anomaly index of the current monitoring window to an exponential function with the natural constant as the base; and a warning index of the current monitoring window is obtained by mapping the product of the relative anomaly index density value of the current monitoring window and the second mapping value to a sigmoid function.

[0075] The calculation model for the early warning indicators in the current monitoring window is as follows:

[0076] ,

[0077] in, This indicates the warning indicator for the current monitoring window a. This represents the relative anomaly index density value of the current monitoring window a. The larger the value, the stronger the persistence or concentration of the abnormal state in the time dimension. This represents the relative anomaly index of the current monitoring window 'a', where 'e' represents the natural constant and the second mapping value. Used to amplify the weight of states with a high degree of abnormality, thereby enhancing the sensitivity to identify potentially severe contact degradation states; sigmoid represents the normalization function, used to map the warning indicators to a preset interval, which facilitates subsequent classification judgment and application;

[0078] By constructing the above-mentioned abnormal early warning indicators, this application can not only reflect the degree of abnormality of the contact quality of the trip switch at the current moment, but also comprehensively characterize the evolution characteristics of the abnormality in the time dimension, and effectively distinguish between short-term disturbances and the actual performance degradation state, thereby improving the stability and reliability of trip switch contact quality monitoring and early warning.

[0079] Finally, after obtaining the early warning indicators for the contact quality of the tripped switch in the current monitoring window 'a' of the target energy meter, the contact status of the tripped switch is judged and warned by setting an early warning threshold T. The early warning threshold T can be a reference threshold pre-set based on the statistical characteristics of historical operating data or maintenance experience. For example, statistical analysis can be performed on the early warning indicators of similar energy meters during long-term stable operation, selecting their mean-weighted standard deviation, quantile threshold, or typical indicator levels before historical faults as reference thresholds; or, combined with on-site maintenance experience, the early warning indicator range corresponding to tripped switches with confirmed poor contact can be used as an abnormal reference range. When the early warning indicator corresponding to the current monitoring window is greater than the early warning threshold T, it is determined that the contact quality of the tripped switch corresponding to the energy meter has an abnormal risk, triggering an early warning and marking the monitoring window as a high-risk window; otherwise, no early warning is triggered. Through this method, the contact quality of tripped switches can be identified in a timely and accurate manner, providing a basis for subsequent maintenance decisions and condition-based maintenance.

[0080] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0081] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A fault online monitoring and analysis energy meter based on multi-source data fusion, characterized in that, include: The data acquisition module is used to acquire the temperature sequence and contact resistance sequence of the trip switch contacts of the target energy meter and other energy meters that are the same as the target energy meter; A monitoring window is set up to divide the temperature and contact resistance sequences. The independent resistance anomaly analysis module is used to obtain the resistance anomaly value of the current monitoring window based on each contact resistance in the current monitoring window of the target energy meter's contact resistance sequence. The relative anomaly analysis module is used to record the monitoring windows corresponding to other energy meters that are the same as the target energy meter as the comparison window; to obtain the comparability evaluation value of the comparison window based on the temperature difference between the current monitoring window and the comparison window; and to obtain the reliability of the comparison result of the comparison window based on the comparability evaluation value of the comparison window and the commissioning time of the energy meter corresponding to the comparison window. The relative abnormality index of the current monitoring window is obtained by using the resistance abnormality value of the current monitoring window, the resistance abnormality value of each comparison window, and the reliability of the comparison results. The early warning monitoring module is used to obtain each reference window of the current monitoring window from the monitoring windows before the current monitoring window of the target energy meter using relative anomaly indicators; to obtain the relative anomaly indicator density value of the current monitoring window using the relative anomaly indicators of the current monitoring window, the relative anomaly indicators of each reference window, and the time distance between each reference window and the current monitoring window; and to fuse the relative anomaly indicator density value and the relative anomaly indicators of the current monitoring window to obtain the early warning indicator. Warnings are issued based on the aforementioned warning indicators; The process of obtaining a comparability assessment value for a comparison window based on the temperature difference between the current monitoring window and a comparison window includes: A first temperature similarity is obtained by negatively correlating the normalized value of the dynamic time-normalized distance between the temperature sequence within the current monitoring window and the temperature sequence within a comparison window using an exponential function with a base of the natural constant. A second temperature similarity is obtained by negatively correlating the normalized value of the absolute value of the difference between the mean temperature within the current monitoring window and the mean temperature within the comparison window using an exponential function with a base of the natural constant. The first temperature similarity and the second temperature similarity are multiplied to obtain the comparability evaluation value of the comparison window. The reliability of the comparison results obtained based on the comparability evaluation value of a comparison window and the commissioning time of the corresponding electricity meter for that comparison window includes: The reliability of the comparison results of a comparison window is obtained by negatively mapping the time distance between the commissioning time of the electricity meter corresponding to a comparison window and the time corresponding to the comparison window using an exponential function with the natural constant as the base, and multiplying it by the comparability evaluation value of the comparison window. The process of obtaining the relative anomaly index of the current monitoring window by utilizing the resistance anomaly value of the current monitoring window, the resistance anomaly values ​​of each comparison window, and the reliability of the comparison results includes: The resistance anomaly value of the current monitoring window is compared with the resistance anomaly value of a comparison window to obtain the resistance anomaly value ratio corresponding to the comparison window; the reliability of the comparison results of each comparison window is used as the weight of the resistance anomaly value ratio corresponding to each comparison window to obtain the weighted average of the resistance anomaly value ratio corresponding to the comparison window to obtain the relative anomaly index of the current monitoring window. The step of obtaining the relative anomaly index density value of the current monitoring window using the relative anomaly indexes of the current monitoring window, the relative anomaly indexes of each reference window, and the time distance between each reference window and the current monitoring window includes: Obtain the normalized value of the absolute value of the difference between the relative anomaly index of the current monitoring window and a reference window, and denot it as the relative anomaly difference corresponding to the reference window; multiply the relative anomaly difference corresponding to the reference window by the normalized value of the time distance between the reference window and the current monitoring window to obtain the relative anomaly difference of fusion time corresponding to the reference window; use an exponential function with the natural constant as the base to perform a negative correlation mapping on the mean of the relative anomaly differences of fusion time corresponding to each reference window to obtain the relative anomaly index density value of the current monitoring window. The method of fusing the relative anomaly index density value and the relative anomaly index of the current monitoring window to obtain the early warning index includes: A second mapping value is obtained by mapping the relative anomaly index of the current monitoring window to an exponential function with the natural constant as the base; the warning index of the current monitoring window is obtained by mapping the product of the relative anomaly index density value of the current monitoring window and the second mapping value to a sigmoid function.

2. The online fault monitoring and analysis energy meter based on multi-source data fusion according to claim 1, characterized in that, The step of obtaining the abnormal resistance value of the current monitoring window based on each contact resistance in the current monitoring window of the target energy meter's contact resistance sequence includes: The coefficient of variation of each contact resistance in the current monitoring window of the contact resistance sequence of the target energy meter is mapped using an exponential function with the natural constant as the base to obtain a first mapped value. The resistance anomaly value of the current monitoring window is obtained by multiplying the first mapped value by the ratio of the mean value of each contact resistance in the current monitoring window of the contact resistance sequence of the target energy meter to the reference value of the contact resistance of the target energy meter.

3. The online fault monitoring and analysis energy meter based on multi-source data fusion according to claim 1, characterized in that, The step of obtaining reference windows from the monitoring windows preceding the current monitoring window of the target energy meter using relative anomaly indicators includes: Obtain a preset number of different monitoring windows from the monitoring windows before the current monitoring window of the target energy meter that have the closest relative abnormality index to the current monitoring window as reference windows for the current monitoring window.

Images