A method and system for early warning of electricity meter faults based on electrothermal coupling characteristics

Abstract

This invention discloses a method and system for early warning of electricity meter faults based on electrothermal coupling characteristics. The method involves acquiring temperature and load signals and performing overheating trigger detection on the electricity meter to determine if a suspected overheating state exists. If a suspected overheating state is found, the electrothermal coupling characteristics and thermal stress accumulation characteristics of the electricity meter are calculated based on the temperature and load signals. The risk level of the electricity meter is then determined based on these characteristics. This invention enables online monitoring and graded quantitative assessment of internal contact degradation, abnormal overheating, and potential burn-out risks in smart electricity meters while maintaining controllable computational complexity. It combines the advantages of physical interpretability and low computational complexity.

Description

Technical Field

[0001] This invention relates to the field of power metering equipment operation status monitoring and intelligent diagnosis technology, specifically to a power meter fault early warning method and system based on electrothermal coupling characteristics. Background Technology

[0002] With the continuous improvement of power distribution network intelligence and the digitalization of power consumption equipment, smart meters have been widely deployed in various power consumption scenarios, undertaking key functions such as power metering, operation monitoring, and power consumption information collection. However, during long-term operation, factors such as load fluctuations, loose wiring, increased contact resistance, changes in ambient temperature, and abnormal operating conditions can easily cause abnormal overheating in critical internal components of the meter. When overheating is not detected and intervened in time, it may further evolve into accelerated contact deterioration, material aging, or even local thermal runaway, ultimately leading to meter burnout, causing metering failure and safety hazards. Extensive engineering practice shows that meter burnout is not an instantaneous failure, but a gradual failure process caused by the long-term accumulation of local overheating. Especially under conditions of contact deterioration, even with small load changes, an abnormal increase in contact resistance can significantly amplify the Joule heating effect, causing a continuous rise in local temperature, which may eventually lead to thermal runaway and meter burnout. Meter burnout not only causes the failure of power metering function but may also cause electrical safety hazards, affecting the safe and stable operation of the power distribution system. Existing monitoring methods for early warning of meter burnout mostly rely on single temperature thresholds, temperature rise rates, or abnormal changes in electrical quantities such as current and voltage for judgment. These methods have the following shortcomings in engineering applications: First, a single temperature or temperature rise indicator is insufficient to distinguish between normal high-load operating conditions and heat accumulation phenomena caused by abnormal contact degradation, easily leading to false alarms; second, relying solely on instantaneous characteristics cannot reflect the long-term cumulative effect of thermal stress, making it difficult to quantitatively assess the risk of meter burnout; third, some methods introduce complex models or high-dimensional features, resulting in high computational complexity, which is not conducive to long-term stable operation in resource-constrained terminals such as electricity meters.

[0003] The existing methods for monitoring overheating of electricity meters have the following problems: (1) The existing early warning methods for overheating or burning of electricity meters mostly use the absolute value of temperature, the rate of temperature rise or a fixed temperature threshold as the main criteria. In actual operation, electricity meters are often in a state of frequent load fluctuation or phased high load operation. It is difficult to accurately distinguish between the load temperature rise under normal operating conditions and the abnormal heat accumulation caused by abnormal operating conditions such as contact deterioration and abnormal increase in contact resistance. It is easy to generate false alarms during high load operation or to miss alarms when the contact gradually deteriorates but the load change is not obvious, thus affecting the accuracy of overheating early warning results and engineering reliability. (2) The failure of electricity meters is usually not a sudden failure, but a gradual failure process caused by the combined effects of local contact deterioration, material aging and long-term accumulation of thermal stress. However, existing methods mostly focus on instantaneous temperature, instantaneous temperature rise rate or short-term characteristic changes, and lack a systematic characterization of the cumulative effect of thermal stress on the time scale. It is difficult to reflect the evolution law of the contact resistance gradually increasing with the extension of the operating time, which makes it impossible to continuously quantify the risk of meter burn-out, and also makes it difficult to achieve graded identification and early warning of different stages. (3) Although some existing technologies introduce multi-dimensional time-frequency features, complex statistical models or deep learning methods to improve diagnostic accuracy, they often rely on high sampling rate data, large-scale feature calculation or model training process, which have high requirements for computing resources, storage resources and data integrity. This is not conducive to long-term stable operation in embedded and resource-constrained terminals such as electricity meters, and the engineering implementation cost is high. It is difficult to meet the comprehensive requirements of real-time performance, stability and maintainability under large-scale deployment conditions. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for early warning of faults in smart energy meters based on electrothermal coupling characteristics, which addresses the above-mentioned problems in the prior art. The present invention can realize online monitoring and graded quantitative assessment of internal contact deterioration, abnormal overheating and potential meter burn-out risk of smart energy meters while ensuring controllable computational complexity. It has the advantages of both physical interpretability and low computational complexity.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A fault early warning method for electricity meters based on electrothermal coupling characteristics includes the following steps: S1: Collect temperature and load signals and perform overheating trigger judgment on the energy meter to determine whether the energy meter is in a suspected overheating state. If the energy meter is in a suspected overheating state, proceed to step S2; otherwise, end and exit. S2, calculate the electrothermal coupling characteristics and thermal stress accumulation characteristics of the energy meter based on the temperature signal and load signal, and determine the risk level of the energy meter based on the electrothermal coupling characteristics and thermal stress accumulation characteristics.

[0006] Optionally, step S1, which involves acquiring the original temperature signal and load signal and performing an overheat trigger judgment on the energy meter to determine whether the energy meter is in a suspected overheating state, includes: S1.1, Acquire temperature and load signals; S1.2, filter the temperature signal and calculate the temperature rise rate characteristics, and calculate the load change rate characteristics based on the load signal; S1.3, determine whether the temperature rise rate characteristic and load change rate characteristic meet the preset suspected overheating condition judgment conditions. If the preset suspected overheating condition judgment conditions are met, the energy meter is determined to be in a suspected overheating state; otherwise, the energy meter is determined to be in normal operation state.

[0007] Optionally, in step S1.2, the function expression for filtering the temperature signal and calculating the temperature rise rate characteristic is as follows: ; ; in, for The characteristics of the rate of temperature rise at any given time. and They are respectively and The temperature signal after time-shifting average filtering The sampling interval is... This is the window size for the moving average filter. This represents the i-th temperature signal within the window of the moving average filter.

[0008] Optionally, in step S1.2, the functional expression for calculating the load change rate characteristic is: ; in, for The characteristics of the load change rate at any given time. and They are respectively and Load signal at any given time The sampling interval is denoted as .

[0009] Optionally, the function expression for the preset suspected overheating state judgment condition in step S1.3 is: and ; in, for The characteristics of the rate of temperature rise at any given time. The threshold value for the temperature rise rate characteristic. for The characteristics of the load change rate at any given time. The threshold value is used to define the load change rate characteristic.

[0010] Optionally, the functional expressions for calculating the electrothermal coupling characteristics and thermal stress accumulation characteristics of the energy meter in step S2 are as follows: ; ; in, The electrothermal coupling characteristics of the electricity meter For the rate of temperature rise, for Temperature signal at any given time Use the historical baseline value of the temperature signal or the average value of a specified sliding window; For load change rate, for Load signal at any given time Use the historical baseline value of the load signal or specify the average value of a sliding window; Characterized by thermal stress accumulation; This represents the number of sampling points within the cumulative time window. This refers to the internal temperature of the electricity meter or the temperature of a specified location at the k-th sampling point within the cumulative time window. This is the temperature reference value inside the electricity meter or at a designated location. ,in For ambient temperature, For the allowable safe temperature rise threshold, For index weights, , The sampling interval is denoted as .

[0011] Optionally, determining the risk level of the electricity meter in step S2 based on its electrothermal coupling characteristics and thermal stress accumulation characteristics includes: S2.1, construct a two-dimensional feature vector from the electrothermal coupling characteristics and thermal stress accumulation characteristics of the electricity meter; S2.2, standardize the two-dimensional feature vector to obtain the standardized two-dimensional feature vector; S2.3 Calculate the distance between the standardized two-dimensional feature vector and the cluster centers obtained by clustering based on historical datasets containing multiple risk levels, and output the risk level corresponding to the cluster center with the smallest distance as the final determined risk level of the electricity meter.

[0012] The present invention also provides a fault early warning system for electricity meters based on electrothermal coupling characteristics, including a microprocessor and a memory interconnected thereto, wherein the microprocessor is programmed or configured to execute the fault early warning method for electricity meters based on electrothermal coupling characteristics.

[0013] The present invention also provides a computer-readable storage medium storing a computer program or instructions that are programmed or configured to execute the energy meter fault early warning method based on electrothermal coupling characteristics by a processor.

[0014] The present invention also provides a computer program product, including a computer program or instructions, which are programmed or configured to execute the energy meter fault early warning method based on electrothermal coupling characteristics via a processor.

[0015] Compared with the prior art, the present invention can mainly achieve the following beneficial effects: (1) The present invention constructs electrothermal coupling features (electrothermal coupling features) to organically combine load change information with temperature response characteristics and introduces load change constraints, effectively overcoming the limitations of traditional methods that rely solely on absolute temperature values ​​or temperature rise rates for judgment. Under complex load fluctuations and phased high load operation conditions, this method can effectively distinguish between load temperature rise and abnormal thermal response amplification caused by contact degradation under normal operating conditions, significantly reducing the false alarm probability caused by load changes, improving the accuracy and robustness of overheat identification, and is suitable for the complex and ever-changing power consumption environment in actual operation. (2) The integral-type thermal stress accumulation feature proposed in the present invention describes the long-term evolution process of abnormal thermal accumulation inside the energy meter on a time scale. By jointly modeling the temperature over-limit amplitude and duration, the quantitative characterization of the abnormal thermal accumulation effect during the long-term operation of the energy meter is realized. Compared with methods that rely solely on instantaneous temperature or short-term features, the present invention can reflect the gradual degradation of contact resistance and the accumulation process of material aging, thereby significantly improving the accuracy of the early warning assessment of meter burnout. (3) This invention adopts a layered early warning mechanism. In the overheating trigger discrimination layer, only low-complexity feature calculations and threshold determinations are performed. The subsequent burn-out early warning evaluation layer is only triggered when an abnormal heat accumulation trend is detected, which effectively reduces the overall computational load and storage requirements. This design takes into account both early warning sensitivity and engineering feasibility, and is suitable for long-term stable operation in electricity meter terminals with limited computing resources, and has good conditions for large-scale deployment. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the basic process of the method in an embodiment of the present invention. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

[0018] like Figure 1 As shown, the energy meter fault early warning method based on electrothermal coupling characteristics in this embodiment includes the following steps: S1: Collect temperature and load signals and perform overheating trigger judgment on the energy meter to determine whether the energy meter is in a suspected overheating state. If the energy meter is in a suspected overheating state, proceed to step S2; otherwise, end and exit. S2, calculate the electrothermal coupling characteristics and thermal stress accumulation characteristics of the energy meter based on the temperature signal and load signal, and determine the risk level of the energy meter based on the electrothermal coupling characteristics and thermal stress accumulation characteristics.

[0019] Step S1 involves acquiring the original temperature signal and load signal, and performing an overheat trigger detection on the energy meter to determine if the energy meter is in a suspected overheating state. This includes: S1.1, Acquire temperature and load signals; S1.2, filter the temperature signal and calculate the temperature rise rate characteristics, and calculate the load change rate characteristics based on the load signal; S1.3, determine whether the temperature rise rate characteristic and load change rate characteristic meet the preset suspected overheating condition judgment conditions. If the preset suspected overheating condition judgment conditions are met, the energy meter is determined to be in a suspected overheating state; otherwise, the energy meter is determined to be in normal operation state.

[0020] Step S1.1 includes placing temperature sensors at key heat-generating parts inside the electricity meter, and simultaneously collecting load current data to obtain temperature sequences T. i With load sequence L i Based on the sampled data, temperature rise rate characteristics and load change rate characteristics are constructed and compared with the judgment thresholds obtained from historical data, serving as the basic input for subsequent overheating trigger judgment and burn-out early warning assessment.

[0021] In step S1.2, the function expression for filtering the temperature signal and calculating the temperature rise rate characteristic is as follows: ; ; in, for The characteristics of the rate of temperature rise at any given time. and They are respectively and The temperature signal after time-shifting average filtering The sampling interval is... This is the window size for the moving average filter. Let be the i-th temperature signal within the window of the moving average filter. The temperature signal is smoothed using a moving average filter, with a window length of . Sliding time window Moving average filtering can suppress measurement noise and short-term random fluctuations. In this embodiment, the length of the sliding time window... This involves performing a moving average and continuous judgment on 10 consecutive sampling points, with a corresponding time length of 10 minutes. This time scale can effectively filter out instantaneous disturbances while maintaining sufficient sensitivity to abnormal heat accumulation.

[0022] In step S1.2, the functional expression for calculating the load change rate characteristic is: ; in, for The characteristics of the load change rate at any given time. and They are respectively and Load signal at any given time The sampling interval is denoted as .

[0023] During overheating trigger detection, based on the feature parameters (temperature rise rate feature and load change rate feature) output in step S1.2, a threshold decision or lightweight classification model can be used to initially determine the operating status of the electricity meter, outputting a normal state or a suspected overheating state. As the first level of the hierarchical early warning structure, the overheating trigger detection layer's core function is to perform preliminary screening of the electricity meter's operating status under low computational complexity. It only triggers the subsequent high-precision burn-out early warning assessment process when an abnormal heat accumulation trend is detected, thereby significantly reducing the overall system computational load while ensuring sensitivity. The logic implementation of this layer is based on the joint analysis of multi-source features, specifically including features based on temperature rise rate. With load change rate characteristics Joint discrimination is performed through collaborative analysis. Select... A preset threshold is set for the temperature rise rate characteristic. This is the threshold for the normal range of load change rate. and The selection is based on the following principles: The temperature rise limit for electricity meters is determined by referring to the specific usage scenario and combining it with historical normal operation data. The threshold is set according to the typical load fluctuation range of the electricity meter to ensure that periodic or transient load changes do not trigger false alarms. Furthermore, the threshold can be dynamically optimized using offline training data to adapt to different installation environments and operating conditions. Under normal high load conditions, and Synchronous changes; in the early stages of contact degradation, even with gradual load changes, the increased contact resistance amplifies the Joule heating effect, leading to... Significantly increased and Keep it at a low level. Therefore, the judgment criterion is set as follows: if... continued The number of sampling periods exceeds the preset threshold. And the same period load change rate If the temperature does not rise significantly and synchronously, it is determined that there is an abnormal thermal response amplification phenomenon in the current state, and "suspected overheating state" is output, immediately triggering the burn-in warning assessment layer, and continuing to step S3. Otherwise, "normal operation state" is output, no subsequent processing is triggered, and the system continues to monitor. Specifically, the function expression of the preset suspected overheating state judgment condition in step S1.3 of this embodiment is: and ; in, for The characteristics of the rate of temperature rise at any given time. The threshold value for the temperature rise rate characteristic. for The characteristics of the load change rate at any given time. This is the threshold for the load change rate characteristic. That is: when... and If the condition is not met, the output will show "suspected overheating state"; otherwise, the output will show "normal operating state". Under complex operating conditions such as short-term high load, periodic load fluctuations, or changes in ambient temperature, this design introduces load change constraints in the initial discrimination stage. It triggers the subsequent burn-in warning assessment process only when contact degradation or abnormal heat accumulation trends occur. This effectively distinguishes between normal high-load temperature rise and abnormal overheating caused by contact degradation. While ensuring sensitivity to early abnormal heat accumulation, it significantly reduces the false alarm rate caused by operating condition fluctuations, providing a highly reliable input for subsequent refined burn-in warning assessment.

[0024] In this embodiment, the constructed temperature rise rate and load change rate are jointly analyzed during overheat trigger detection. In this example, based on historical statistical data under normal operating conditions, the temperature rise rate of key components of the electricity meter has been stably distributed at 0.005 for a long period of time. 0.025℃ / min. When contact degradation occurs, the temperature rise rate increases significantly, typically exceeding 0.04℃ / min. Considering the distribution of normal and abnormal operating conditions, the temperature rise rate threshold is set as follows in this embodiment: ; When the continuous sampling window satisfies At that time, it was considered that the rate of temperature rise had exceeded the normal operating range.

[0025] According to historical operational data, the rate of change of normal electrical load over a 1-minute timescale typically does not exceed 0.05I. n / min. To avoid misjudgment caused by periodic load fluctuations or transient shocks, the load change rate threshold is set to: ; When satisfied At that time, it is considered that the current load change is within the normal or gradual range.

[0026] Within a consecutive N=10 sampling points, if the following conditions are met simultaneously: ; If the current energy meter is found to have an abnormal thermal response amplification phenomenon, it will output a suspected overheating state and trigger the meter burn-out warning assessment layer; otherwise, it will be determined to be in normal operating state.

[0027] Selecting a set of typical experimental data, under normal contact conditions, the load is 0.8I. n Increased to 1.1I n The temperature of a critical component inside the electricity meter rose from 32.0℃ to 32.6℃ over a period of 30 minutes. Therefore, its average temperature rise rate is: ; The average load change rate is: ； A reasonable correlation is maintained between temperature changes and load changes, indicating normal operation.

[0028] Under simulated contact degradation conditions and the same load variation, the temperature of key internal components of the electricity meter rose from 33.0℃ to 41.8℃, with the average temperature rise rate being: ; The load change rate is still approximately At this point, the rate of temperature rise increases significantly and is mismatched with the load change, triggering the overheating detection condition and proceeding to step S3.

[0029] When the judgment result indicates a suspected overheating state, a burn-out warning assessment is triggered, further calculating the electrothermal coupling characteristics and thermal stress accumulation characteristics. This embodiment proposes an electrothermal coupling characteristic to characterize the relationship between load changes and temperature response of an energy meter, reflecting the abnormal thermal sensitivity of the energy meter under load disturbance conditions. This characteristic is defined as the ratio of the energy meter's temperature rise rate to the load change rate, and a normalization process is introduced to reduce the impact of differences in operating conditions. The original electrothermal coupling characteristics inside the energy meter are defined. for: ; in, This refers to the temperature inside or in critical parts of the electricity meter. Indicates the rate of temperature rise; Electrical quantities characterizing load conditions. This represents the load change rate. Based on historical operating data statistics, it shows the original electrothermal coupling characteristics under different operating conditions. The typical range of values ​​is shown in Table 1.

[0030] Table 1: Original electrothermal coupling characteristics under different operating conditions Typical range of values

[0031] The interval is not used as a fixed threshold, but rather as the physical interpretation basis for the clustering results, ensuring the method's adaptability and engineering robustness. When the electrothermal coupling characteristics increase significantly, it indicates that the increased contact resistance leads to a significant amplification of the thermal response under load disturbances. According to experimental data, under normal contact conditions, the original electrothermal coupling characteristics remain. for: ; Under contact degradation conditions, the original electrothermal coupling characteristics are present. for: ; As can be seen, the results are consistent with the experimental conclusions.

[0032] To enhance robustness, a normalized form is introduced: ; in, The electrothermal coupling characteristics of the electricity meter For the rate of temperature rise, for Temperature signal at any given time Use the historical baseline value of the temperature signal or the average value of a specified sliding window; For load change rate, for Load signal at any given time This refers to the historical reference value of the load signal or the average value within a specified sliding window. Under normal operating conditions, there is a significant lag between load changes and temperature changes, the temperature rise rate is small, and electrothermal coupling characteristics are present. Low. Under adverse chemical conditions, contact resistance increases, and even with small load changes, the temperature rise rate increases significantly, resulting in a thermal response amplification effect and electrothermal coupling characteristics. Significantly increased. This feature characterizes the abnormal thermal sensitivity caused by load disturbances, and compared with a single temperature rise feature, it effectively distinguishes abnormal heat accumulation by incorporating load change information.

[0033] Electricity meter burnout is not an instantaneous failure, but rather a result of long-term accumulated thermal stress leading to contact degradation, which in turn causes thermal runaway and even meter burnout. Instantaneous temperature or temperature rise rate only reflects the current state; however, in actual operating conditions, contact resistance increases with accumulated thermal stress, and the greater the thermal stress and the longer its duration, the faster the degradation rate. Addressing the long-term accumulation of thermal stress in electricity meter burnout, this invention proposes an integral thermal stress accumulation characteristic. This characteristic is derived by integrating the amplitude and duration of the temperature exceeding a reference value during meter operation to characterize the cumulative degree of contact degradation and material aging within the meter. Therefore, the original thermal stress accumulation characteristic can be defined as: ; in, Original thermal stress accumulation characteristics, This represents the number of sampling points within the cumulative time window. To obtain the maximum value, This refers to the internal temperature of the electricity meter or the temperature of a specified location at the k-th sampling point within the cumulative time window. This is the temperature reference value inside the electricity meter or at a designated location. ,in For ambient temperature, For the allowable safe temperature rise threshold, The sampling interval is specified. In this embodiment, the ambient temperature is... It is 23℃. ,

[0034] To more accurately reflect the nonlinear characteristics of material aging and contact degradation, an exponential weight is introduced. ,and If the value is greater than 1, then the functional expression for the thermal stress accumulation characteristic can be obtained: ; Characterized by thermal stress accumulation; This represents the number of sampling points within the cumulative time window. This refers to the internal temperature of the electricity meter or the temperature of a specified location at the k-th sampling point within the cumulative time window. This is the temperature reference value inside the electricity meter or at a designated location. ,in For ambient temperature, For the allowable safe temperature rise threshold, For index weights, The sampling interval. When When it is a linear accumulation, when When the value is >1, the weight of the high-temperature segment is significantly amplified. In this embodiment, we take... Under contact degradation conditions, the duration of temperature exceeding the baseline value is approximately 48 minutes, and the average over-limit temperature is approximately 4.2℃. Therefore, the cumulative thermal stress is approximately... .

[0035] By introducing nonlinear weights, higher weights can be assigned to the high-temperature range to more accurately reflect the nonlinear evolution characteristics of material aging and contact degradation. Therefore, the functional expressions for calculating the electrothermal coupling characteristics and thermal stress accumulation characteristics of the energy meter in step S2 of this embodiment are as follows: ; ; Characterized by thermal stress accumulation; This represents the number of sampling points within the cumulative time window. This refers to the internal temperature of the electricity meter or the temperature of a specified location at the k-th sampling point within the cumulative time window. This is the temperature reference value inside the electricity meter or at a designated location. ,in For ambient temperature, For the allowable safe temperature rise threshold, For index weights, The sampling interval is defined as follows. This functional expression proposes an integral feature to characterize the degree of thermal stress accumulation in electricity meters. By accumulating the amplitude and duration of temperature exceeding a reference value during meter operation, it characterizes the degree of progressive damage caused by abnormal heat accumulation during long-term operation, thereby achieving a quantitative assessment of the meter burnout warning level. Compared with diagnostic methods based solely on instantaneous temperature or temperature rise rate, the thermal stress accumulation feature described in this invention can effectively reflect the cumulative effects of internal contact degradation and material aging under long-term operating conditions, improving the accuracy of meter burnout warning assessment.

[0036] In step S2, determining the risk level of the electricity meter based on its electrothermal coupling characteristics and thermal stress accumulation characteristics can be achieved using methods such as lookup tables, machine learning models, and clustering. For example, as an optional implementation, in this embodiment, step S2, determining the risk level of the electricity meter based on its electrothermal coupling characteristics and thermal stress accumulation characteristics includes: S2.1, construct a two-dimensional feature vector from the electrothermal coupling characteristics and thermal stress accumulation characteristics of the electricity meter; S2.2, the two-dimensional feature vector is standardized to obtain a standardized two-dimensional feature vector; for example, as an optional implementation method, in order to eliminate the influence of differences in feature scale and numerical range on the clustering results, this embodiment adopts Z-score standardization of the two-dimensional feature vector to achieve standardization; S2.3 Calculate the distance between the standardized two-dimensional feature vector and the cluster centers obtained by clustering based on historical datasets containing multiple risk levels. The risk level corresponding to the cluster center with the smallest distance is then used as the final determined risk level output for the electricity meter. This can be expressed as: ; in, To determine the final risk level of the electricity meter, The standardized two-dimensional feature vector, This represents the k-th cluster center. The risk level can be set as needed. For example, as an optional implementation, the risk levels in this embodiment include levels I to III, defined as follows: Level I: Potential overheating, electrothermal coupling characteristics are slightly higher than normal, and thermal stress accumulation is at a low level; Level II: Significant overheating, with a marked increase in electrothermal coupling characteristics, and thermal stress accumulation has entered an accelerated phase; Level III: High risk of burn-in, electrothermal coupling characteristics remain high, thermal stress accumulation has approached or exceeded safety limits, and there is a risk of thermal runaway.

[0037] In this embodiment, the pre-clustering based on a historical dataset containing data on multiple risk levels includes: S101, construct two-dimensional feature vectors for each sample in the historical dataset: ; in, Let be the two-dimensional feature vector of the i-th sample, and in the superscript . This is a transpose operation; S102, to eliminate the influence of differences in feature scale and numerical range on clustering results, Z-score standardization is performed on the two-dimensional feature vectors of each sample in the historical dataset: ; in, The standardized result of the j-th dimension feature of the i-th sample. Let j be the original value of the j-th dimension feature of the i-th sample. Let be the mean of the j-th feature in the historical dataset. Let be the standard deviation of the j-th feature; finally, we obtain the standardized two-dimensional feature vector of the i-th sample. ; S103, Perform K-means clustering on the standardized two-dimensional feature vectors of each sample in the historical dataset, and input the standardized two-dimensional feature vector matrix. , ~ These are the standardized two-dimensional feature vectors of samples 1 to N, where N is the number of samples in the historical dataset; the number of clusters in K-means clustering is K=3, the maximum number of iterations is M, and the convergence threshold is... Finally, K-means clustering is used to divide and output the warning level label for each sample. and various types of cluster centers .

[0038] Step S103 specifically includes: S201, randomly select K samples from the historical dataset as initial cluster centers: ; in, Let the initial value be the cluster center of the k-th category. The standardized two-dimensional feature vector of the k-th randomly selected sample from the historical dataset; the number of iterations t is initialized to 0; S202, for each sample in the historical dataset, calculate the Euclidean distance between its standardized two-dimensional feature vector and each cluster center: ; in, The standardized two-dimensional feature vector of the i-th sample To the k-th cluster center in the t-th iteration The Euclidean distance between them; S203, assign the sample to the nearest cluster: ; in, The cluster assigned to the i-th sample in the t-th iteration; finally, the cluster at the t-th iteration is obtained: ; in, This refers to the k-th cluster at the t-th iteration. S204, for each cluster in the t-th iteration, recalculate its cluster center: ; in, Let K be the cluster center of the k-th cluster at the (t+1)-th iteration. The number of samples in the k-th cluster at the t-th iteration; S205, Termination condition for determining the change in cluster centers: ; in, Let K be the cluster center of the k-th cluster in the t-th iteration. This is the convergence threshold; S206: If the termination condition is met or the iteration count t equals the preset maximum iteration count M, then the K-means clustering is considered complete, the process ends, and the function returns. Otherwise, the iteration count is incremented by 1, and the process jumps to step S202 to continue iterating. After clustering is complete, each sample receives its final label. Output clustering results based on labels. Corresponding warning levels: Corresponding to Level I potential overheating, Corresponding to Level II obvious overheating, This corresponds to Level III high-risk burn-in.

[0039] To verify the effectiveness of the meter fault early warning method based on electrothermal coupling characteristics in this embodiment, the electrothermal coupling characteristics and thermal stress accumulation characteristics obtained in this embodiment are used to construct a feature vector, which is then used as the input to the meter burnout early warning assessment layer. Under the above-mentioned judgment parameter settings, historical operating data collected in the experimental platform are statistically analyzed. The operating status is classified using the K-means clustering algorithm, and manually labeled contact status and temperature anomaly records are used as references to verify the overheating trigger discrimination results. Statistical results show that normal contact status samples are stably distributed in the Level I potential overheating area, contact deterioration status samples are mainly distributed in the Level II obvious overheating area, and some samples with continuous high temperature are further classified into the Level III high-risk meter burnout area. In the selected sample set, the method of this invention achieves an accuracy rate of 94.6% in identifying abnormal overheating states, with a false alarm rate controlled within 3.1%. It can be seen that the overheating trigger criterion of the meter fault early warning method based on electrothermal coupling characteristics in this embodiment ensures high recognition sensitivity while possessing good stability and engineering robustness, and can provide reliable anomaly screening input for subsequent meter burnout early warning assessment. In this specific embodiment, an overheating test platform for electricity meters is constructed to collect operating data of the electricity meters under different loads and contact conditions. The operating status is initially screened through the overheating trigger discrimination layer. When a suspected overheating state is detected, the electrothermal coupling characteristics and thermal stress accumulation characteristics are further calculated, and the meter burnout warning level is graded based on a preset threshold or clustering results. Experimental results show that the electricity meter fault warning method based on electrothermal coupling characteristics in this embodiment can effectively distinguish between normal operating conditions and abnormal overheating conditions, and accurately identify different levels of meter burnout warnings, demonstrating good engineering application value.

[0040] In summary, the fault early warning method for electricity meters based on electrothermal coupling characteristics in this embodiment significantly reduces computational complexity while ensuring early warning accuracy through hierarchical processing and cluster analysis. It is suitable for long-term, stable online monitoring and hierarchical early warning in resource-constrained terminals such as electricity meters.

[0041] Those skilled in the art will understand that the technical solutions provided by this invention can take the form of a method, system, or computer program product. Therefore, this invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. For example, this invention can provide a power meter fault early warning system based on electrothermal coupling characteristics, including a microprocessor and a memory interconnected, the microprocessor being programmed or configured to execute the power meter fault early warning method based on electrothermal coupling characteristics. This invention can provide a computer-readable storage medium storing a computer program or instructions programmed or configured to execute the power meter fault early warning method based on electrothermal coupling characteristics via a processor. This invention can provide a computer program product including a computer program or instructions programmed or configured to execute the power meter fault early warning method based on electrothermal coupling characteristics via a processor. Furthermore, this invention can take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0042] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for early warning of electricity meter faults based on electrothermal coupling characteristics, characterized in that, Includes the following steps: S1: Collect temperature and load signals and perform overheating trigger judgment on the energy meter to determine whether the energy meter is in a suspected overheating state. If the energy meter is in a suspected overheating state, proceed to step S2; otherwise, end and exit. S2, calculate the electrothermal coupling characteristics and thermal stress accumulation characteristics of the energy meter based on the temperature signal and load signal, and determine the risk level of the energy meter based on the electrothermal coupling characteristics and thermal stress accumulation characteristics.

2. The method for early warning of electricity meter faults based on electrothermal coupling characteristics according to claim 1, characterized in that, Step S1 involves acquiring the original temperature signal and load signal, and performing an overheat trigger detection on the energy meter to determine if the energy meter is in a suspected overheating state. This includes: S1.1, Acquire temperature and load signals; S1.2, filter the temperature signal and calculate the temperature rise rate characteristics, and calculate the load change rate characteristics based on the load signal; S1.3, determine whether the temperature rise rate characteristic and load change rate characteristic meet the preset suspected overheating condition judgment conditions. If the preset suspected overheating condition judgment conditions are met, the energy meter is determined to be in a suspected overheating state; otherwise, the energy meter is determined to be in normal operation state.

3. The method for early warning of electricity meter faults based on electrothermal coupling characteristics according to claim 2, characterized in that, In step S1.2, the function expression for filtering the temperature signal and calculating the temperature rise rate characteristic is as follows: ； ； in, for The characteristics of the rate of temperature rise at any given time. and They are respectively and The temperature signal after time-shifting average filtering The sampling interval is... This is the window size for the moving average filter. This represents the i-th temperature signal within the window of the moving average filter.

4. The method for early warning of electricity meter faults based on electrothermal coupling characteristics according to claim 2, characterized in that, In step S1.2, the functional expression for calculating the load change rate characteristic is: ； in, for The characteristics of the load change rate at any given time. and They are respectively and Load signal at any given time The sampling interval is denoted as .

5. The method for early warning of electricity meter faults based on electrothermal coupling characteristics according to claim 2, characterized in that, The function expression for the preset suspected overheating state judgment condition in step S1.3 is: and ; in, for The characteristics of the rate of temperature rise at any given time. The threshold value for the temperature rise rate characteristic. for The characteristics of the load change rate at any given time. The threshold value is used to define the load change rate characteristic.

6. The method for early warning of electricity meter faults based on electrothermal coupling characteristics according to claim 1, characterized in that, The functional expressions for calculating the electrothermal coupling characteristics and thermal stress accumulation characteristics of the energy meter in step S2 are as follows: ； ； in, The electrothermal coupling characteristics of the electricity meter For the rate of temperature rise, for Temperature signal at any given time Use the historical baseline value of the temperature signal or the average value of a specified sliding window; For load change rate, for Load signal at any given time Use the historical baseline value of the load signal or specify the average value of a sliding window; Characterized by thermal stress accumulation; This represents the number of sampling points within the cumulative time window. This refers to the internal temperature of the electricity meter or the temperature of a specified location at the k-th sampling point within the cumulative time window. This is the temperature reference value inside the electricity meter or at a designated location. ,in For ambient temperature, For the allowable safe temperature rise threshold, For index weights, , The sampling interval is denoted as .

7. The method for early warning of electricity meter faults based on electrothermal coupling characteristics according to claim 1, characterized in that, Step S2, which determines the risk level of the electricity meter based on its electrothermal coupling characteristics and thermal stress accumulation characteristics, includes: S2.1, construct a two-dimensional feature vector from the electrothermal coupling characteristics and thermal stress accumulation characteristics of the electricity meter; S2.2, standardize the two-dimensional feature vector to obtain the standardized two-dimensional feature vector; S2.3 Calculate the distance between the standardized two-dimensional feature vector and the cluster centers obtained by clustering based on historical datasets containing multiple risk levels, and output the risk level corresponding to the cluster center with the smallest distance as the final determined risk level of the electricity meter.

8. A fault early warning system for electricity meters based on electrothermal coupling characteristics, comprising a microprocessor and a memory interconnected, characterized in that, The microprocessor is programmed or configured to execute the energy meter fault early warning method based on electrothermal coupling characteristics as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program or instructions, characterized in that, The computer program or instructions are programmed or configured to execute, via a processor, the method for early warning of electricity meter faults based on electrothermal coupling characteristics as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program or instructions, characterized in that, The computer program or instructions are programmed or configured to execute, via a processor, the method for early warning of electricity meter faults based on electrothermal coupling characteristics as described in any one of claims 1 to 7.

Images