Electric power loop inspection method based on intelligent electric energy meter

Abstract

The invention relates to a power loop inspection method based on an intelligent electric energy meter. The method comprises the following steps: determining a working condition grade according to a current load current effective value, and determining a first threshold and a second threshold from a threshold mapping relation; obtaining a voltage loop self-monitoring error variation, a current loop self-monitoring error variation and a terminal temperature variation of each terminal; performing working condition interference correction on the variation and performing variation trend analysis; determining that there is a loop abnormality in response to at least one of the corrected error variations exceeding a first threshold and at least one of the corrected terminal temperature variations exceeding a second threshold, or in response to a variation trend analysis that there is an early degraded terminal; and determining a fault type according to the combination of the corrected dual-channel error variation and the temperature variation of each terminal. According to the invention, correlation analysis is carried out on the self-monitoring error and the terminal temperature under the background of a unified load working condition, so that the accuracy of power loop abnormity determination under a complex working condition is improved.

Description

Technical Field

[0001] This application relates to the field of smart meters, and in particular to a method for inspecting power circuits based on smart meters. Background Technology

[0002] Smart meters play a core role in electricity metering within power systems. During long-term operation, the metering circuit of a smart meter can degrade in performance due to component aging, environmental changes, and human factors. For example, drift in sampling device parameters can increase metering errors, or oxidation or loosening of the terminal block can increase contact resistance and cause abnormal heating during current flow.

[0003] To monitor the health of the metering circuit during operation, two types of online monitoring methods have emerged in the prior art. One type is metering error monitoring based on self-monitoring signal injection, which injects interharmonic signals into the sampling circuit through a reference source built into the metering chip and extracts the changes in this signal to determine whether an abnormality has occurred in the sampling circuit. The other type is terminal block temperature monitoring, which collects temperature data of each terminal through temperature sensors to determine whether the terminal contact status is normal.

[0004] However, in existing solutions, these two methods operate independently, each with its own judgment threshold and alarm trigger. This independent operation makes it difficult for existing solutions to distinguish between genuine loop faults and parameter fluctuations caused by non-fault reasons. Specifically, self-monitoring errors are prone to jumps due to dynamic load spectrum interference during load changes, and terminal temperatures also rise under normal high-current operating conditions or when ambient temperature fluctuates. The abnormal parameters observed in either case do not necessarily indicate a loop fault. Because there is a lack of correlation analysis between the two types of monitoring results, judging based solely on either type of parameter cannot distinguish between abnormalities caused by loop faults and normal fluctuations caused by changes in operating conditions. When load spectrum interference causes self-monitoring errors to jump, independent error monitoring cannot use the normal state of terminal temperature to rule out false alarms; when high-current operating conditions cause terminal temperature to rise, independent temperature monitoring cannot use the normal state of self-monitoring error to confirm that the temperature rise is within the normal range. The complementary information contained in the two types of monitoring data is not utilized, resulting in insufficient reliability of the judgment results under complex operating conditions. Summary of the Invention

[0005] In order to accurately determine whether there is an abnormality in the power circuit by correlating the self-monitoring error with the terminal temperature under a unified load condition, this application provides a power circuit inspection method based on a smart energy meter.

[0006] This application provides a power circuit inspection method based on a smart energy meter, which adopts the following technical solution: A power circuit inspection method based on a smart energy meter includes: S1. Obtain the current effective value of the load current, determine the current operating condition level based on the ratio of the current effective value of the load current to the rated current, and determine the first threshold and the second threshold from the preset threshold mapping relationship based on the current operating condition level; S2. Obtain the change in self-monitoring error of the voltage loop and the change in self-monitoring error of the current loop in the metering circuit; S3. Obtain the terminal temperature change of each terminal, wherein each terminal includes a current terminal and a voltage terminal; S4. Perform operating condition interference correction on the voltage loop self-monitoring error change, the current loop self-monitoring error change, and the terminal temperature change of each terminal to obtain the corrected voltage loop self-monitoring error change, the corrected current loop self-monitoring error change, and the corrected terminal temperature change of each terminal, and perform trend analysis on the terminal temperature change of each terminal. S5. In response to at least one of the corrected voltage loop self-monitoring error change and the corrected current loop self-monitoring error change exceeding the first threshold and at least one of the corrected terminal temperature changes of each terminal exceeding the second threshold, or in response to determining, based on the trend analysis, that there is a terminal in an early deterioration state among the terminals, it is determined that there is a loop anomaly in the power circuit of the smart energy meter. S6. In response to determining that the power circuit has the circuit abnormality, the fault type is determined based on the combination of the corrected voltage circuit self-monitoring error change, the current circuit self-monitoring error change, and the terminal temperature change of each terminal, specifically the terminal temperature change of the current terminal and the terminal temperature change of the voltage terminal.

[0007] By adopting the above technical solution, the monitoring data of self-monitoring error and terminal temperature are correlated and analyzed under the same load condition. This ensures that jumps in self-monitoring error caused by load spectrum interference will not trigger false alarms when terminal temperature is not synchronously abnormal, and deviations in terminal temperature caused by normal high-current operation or ambient temperature fluctuations will not trigger false alarms when self-monitoring error is not synchronously abnormal. Abnormalities in both parameters must occur synchronously in time to constitute a loop anomaly determination condition, thus suppressing false alarms caused by non-fault factors without reducing the sensitivity of single-parameter detection. Simultaneously, the first and second thresholds are dynamically determined according to the current operating condition level, allowing the joint determination conditions to automatically adjust with the load level, avoiding the applicability contradiction of fixed thresholds between light and heavy load conditions. After an anomaly is detected, the fault type is determined by a four-parameter combination of the dual-channel error change and the temperature change of each terminal, enabling maintenance personnel to address the root cause of the fault without having to check each link individually.

[0008] Optionally, step S6 includes the following sub-steps: S61. In response to the fact that the change in the self-monitoring error of the corrected current loop exceeds the first threshold and the change in the terminal temperature of the current terminal in the terminal temperature change of each corrected terminal exceeds the second threshold, the fault type is determined to be abnormal contact of the current terminal. S62. In response to the fact that the change in the self-monitoring error of the corrected voltage loop exceeds the first threshold and the change in the terminal temperature of the voltage terminal among the changes in the terminal temperature of each corrected terminal exceeds the second threshold, the fault type is determined to be abnormal voltage terminal contact. S63. In response to the fact that the change in the self-monitoring error of the corrected current loop exceeds the first threshold and the change in the terminal temperature of the current terminal among the changes in the terminal temperature of each corrected terminal does not exceed the second threshold, the fault type is determined to be current sampling device degradation. S64. In response to the fact that the change in the self-monitoring error of the corrected voltage loop exceeds the first threshold and the change in the terminal temperature of the voltage terminal among the changes in the terminal temperature of each corrected terminal does not exceed the second threshold, the fault type is determined to be voltage sampling device degradation. S65. In response to the fact that both the corrected change in the self-monitoring error of the current loop and the corrected change in the self-monitoring error of the voltage loop exceed the first threshold and the corrected change in the terminal temperature of each terminal does not exceed the second threshold, the fault type is determined to be metering chip reference drift.

[0009] By employing the above technical solution, the fault type determination is based on the physical causal relationship between the self-monitoring error channel attribution and the terminal temperature response. An increase in terminal contact resistance simultaneously leads to an increase in the self-monitoring error of the corresponding circuit and a rise in the terminal temperature; therefore, the co-occurrence of error anomalies and temperature anomalies on the same side points to a terminal contact anomaly. Meanwhile, drift in the internal parameters of the sampling device only affects the transfer function on the signal path and does not generate additional Joule heat at the terminal contact surface; therefore, an error anomaly but a normal temperature points to degradation of the sampling device. When both channels exhibit simultaneous errors but all terminal temperatures are normal, external circuit causes are ruled out, locking the fault to a drift in the reference source within the metering chip. The five branches exhaustively enumerate all valid combinations of error channels and temperature states, making the determination of each fault type exclusive.

[0010] Optionally, the trend analysis of the terminal temperature change of each terminal performed in S4 includes: S41. Calculate the terminal temperature rise rate between two adjacent inspection cycles; S42. Calculate the terminal temperature difference between each terminal and the other terminals. In response to the terminal temperature rise rate of a terminal exceeding a third threshold and the terminal temperature difference between the terminal and the other terminals exceeding a fourth threshold, determine that the terminal is in an early deterioration state.

[0011] By employing the above technical solution, terminals in the early deterioration stage can be identified through joint detection of two dimensions: temperature rise rate and inter-terminal temperature difference, before the absolute value of the terminal temperature reaches the second threshold. The temperature rise rate reflects the trend of terminal contact resistance deterioration over time, while the temperature difference reflects the degree of temperature deviation of the terminal relative to other terminals within the same meter. Both dimensions must be met simultaneously to confirm early deterioration. For example, a sudden rate jump caused by sensor noise is detected, but the temperature difference is normal; or an ambient temperature gradient is detected, resulting in an inter-terminal temperature difference, but the temperature rise rate of each terminal is consistent. In this technical solution, these two situations will not be misjudged as deterioration signals, thus shifting the fault detection window from a reactive alarm after the terminal temperature exceeds the limit to the early stage when the terminal contact resistance begins to deteriorate.

[0012] Optionally, S41 includes the following sub-steps: S411. Obtain the terminal temperature acquisition value of each terminal in the Kth inspection cycle and the terminal temperature acquisition value in the (K-1)th inspection cycle, where K is a positive integer greater than 1; S412. Divide the difference between the terminal temperature acquisition value of the Kth inspection cycle and the terminal temperature acquisition value of the (K-1)th inspection cycle by the time interval between two adjacent inspection cycles to obtain the terminal temperature rise rate. S413. Perform a sliding window average on the N terminal temperature rise rates calculated over N consecutive inspection cycles to obtain a smoothed terminal temperature rise rate, and compare the smoothed terminal temperature rise rate with the third threshold, where N is a positive integer greater than or equal to 2.

[0013] By employing the above technical solution, a sliding window averaging is performed on the terminal temperature rise rate calculated over multiple consecutive inspection cycles. This suppresses cycle-by-cycle rate fluctuations caused by temperature sensor sampling noise and quantization errors, making the smoothed rate value closer to the steady-state temperature rise rate corresponding to the actual deterioration rate of the terminal contact resistance. The length of the sliding window is controlled by the parameter N. A larger N results in a stronger smoothing effect but a slower response to sudden changes, while a smaller N results in a faster response but a weaker noise immunity. The window can be configured according to the time interval of the inspection cycle and the accuracy level of the temperature sensor.

[0014] Optionally, the operating condition disturbance correction in S4 includes at least one of the following: In response to the change in the effective value of the current load current exceeding a preset proportion of the rated current within a single inspection cycle, a mark is added to the inspection cycle as a load transition period, and the determination that the terminal temperature rise rate exceeds the third threshold is suspended during the load transition period. Before determining that the terminal temperature difference exceeds the fourth threshold, the average value of the terminal temperature change of each terminal is calculated, and the differential temperature change is obtained by subtracting the average value from the terminal temperature change of each terminal. The fourth threshold is then determined based on the differential temperature change instead of the terminal temperature difference. In response to the total harmonic distortion rate of the grid voltage collected by the smart energy meter exceeding the harmonic interference threshold, the change in the self-monitoring error of the voltage loop and the change in the self-monitoring error of the current loop are smoothed by taking the median value through a sliding window before the determination in S5 is executed.

[0015] By employing the above technical solutions, corrections are applied to the impact of three operating conditions—load surges, ambient temperature common-mode interference, and power grid harmonic interference—on early degradation assessment and joint assessment. During load surges, the temperature rise rate of all terminals increases synchronously; pausing the rate assessment avoids misinterpreting normal load responses as degradation signals. The differential-mode temperature change is reduced by subtracting the average temperature change of all terminals to eliminate interference from overall ambient temperature fluctuations on inter-terminal temperature difference comparisons, ensuring that temperature difference assessment only responds to the differential temperature rise between terminals. When the harmonic distortion rate exceeds the limit, the self-monitoring signal extraction is affected by power grid spectrum interference; the error change is smoothed by taking the median value to eliminate the influence of outliers before entering the joint assessment. These three correction methods protect different parameter dimensions in the assessment process, ensuring that joint assessment and early degradation assessment maintain accuracy even under complex operating conditions.

[0016] Optionally, the step of performing the determination of the fourth threshold based on the differential mode temperature change instead of the terminal temperature difference value includes the following sub-steps: S71. During each inspection cycle, acquire the terminal temperature data of each terminal; S72. Calculate the difference between the terminal temperature acquisition value of each terminal and the terminal temperature acquisition value of the previous inspection cycle to obtain the terminal temperature change of each terminal. S73. Calculate the average value of the terminal temperature change of each terminal to obtain the common mode temperature change component; S74. Subtract the common-mode temperature change component from the terminal temperature change of each terminal to obtain the differential-mode temperature change of each terminal; S75. Calculate the difference between the differential mode temperature change of each terminal and the differential mode temperature change of the other terminals, and determine the temperature difference abnormality of the terminal in response to the difference exceeding the fourth threshold.

[0017] By adopting the above technical solution, the calculation process of differential mode temperature change is limited as follows: first, the difference between the temperature acquisition value of the current cycle and the previous cycle of each terminal is obtained as the temperature change; then, the average value of the temperature change of all terminals is calculated to obtain the common mode temperature change component; then, the common mode component is subtracted from the temperature change of each terminal to obtain the differential mode temperature change; finally, the pairwise difference between the differential mode values ​​of each terminal is compared with a fourth threshold. The common mode component carries the ambient temperature change experienced by all terminals in the same meter. After subtracting it, the remaining differential mode component only reflects the independent temperature behavior of each terminal, so that the temperature difference comparison result is not affected by the overall rise and fall of the ambient temperature. Temperature difference abnormality is only triggered when a terminal produces an independent temperature rise different from the other terminals due to increased contact resistance.

[0018] Optionally, marking the inspection cycle as a load transition period further includes: During the load transition period, the terminal temperature data of each terminal is continuously collected and stored as a transition period temperature sequence. In response to the fact that the change in the effective value of the current load current does not exceed the preset ratio of the rated current for M consecutive inspection cycles, the load transition period marker is removed, where M is a positive integer greater than 1; After removing the mark of the load transition period, the expected cumulative temperature rise value of each terminal during the load transition period is calculated according to the current operating condition level corresponding to each inspection cycle during the load transition period. The actual cumulative temperature rise value of each terminal in the transition period temperature sequence is compared with the expected cumulative temperature rise value. In response to the fact that the actual cumulative temperature rise of a terminal exceeds the sum of the expected cumulative temperature rise and a preset deviation, it is determined that the terminal has a superimposed degradation signal during the load transition period, and the terminal with the superimposed degradation signal is determined to be in the early degradation state.

[0019] By adopting the above technical solution, the monitoring blind spot problem caused by the pause rate determination during the load transition period is solved. During the transition period, terminal temperature data is continuously cached. After the transition period ends, the expected temperature rise is accumulated cycle by cycle according to the operating condition level corresponding to each inspection cycle during that period. The actual accumulated temperature rise is then compared with the expected accumulated temperature rise. The actual accumulated temperature rise of normal terminals is determined by the ohmic thermal effect caused by load changes, which matches the expected value calculated based on standard contact resistance. However, for deteriorated terminals where contact resistance has begun to increase, the actual accumulated temperature rise will systematically exceed the expected value. By setting the deviation as a judgment threshold, the deterioration signal that was blocked by the pause determination during the transition period is re-detected after the transition period ends, and the terminal is identified as an early deterioration state to trigger subsequent fault type determination.

[0020] Optionally, the smoothing of the changes in the self-monitoring error of the voltage loop and the self-monitoring error of the current loop using a sliding window median method includes: The harmonic interference level is determined based on the magnitude of the total harmonic distortion rate of the power grid voltage. The width of the sliding window is selected from a preset window width mapping relationship according to the harmonic interference level, wherein the higher the harmonic interference level, the larger the width of the sliding window. In response to the total harmonic distortion rate of the grid voltage exceeding the extreme interference threshold, the changes in the self-monitoring error of the voltage loop and the self-monitoring error of the current loop collected in the corresponding inspection cycle are marked as invalid data, and the invalid data is not included in the median calculation of the sliding window.

[0021] By employing the above technical solution, the sliding window width is automatically adjusted according to the magnitude of the harmonic distortion rate. Under mild harmonic interference, the window is narrower to preserve the timeliness of the error data; under moderate harmonic interference, the window is wider to enhance the smoothing effect. When the harmonic distortion rate exceeds the extreme interference threshold, the error data for the corresponding period is directly marked as invalid and excluded from the window. Invalid data does not participate in the median calculation, avoiding the problem of the median itself deviating from the true error trend due to a large amount of distorted data filling the window. The tiered response allows the smoothing strategy to automatically adjust the trade-off between timeliness and reliability according to the interference intensity, rather than being fixed at a certain compromise point.

[0022] Optional, also includes: If the load transition period marker and the total harmonic distortion rate of the grid voltage exceeding the harmonic interference threshold are both established within the same inspection cycle, it is determined that the inspection cycle is in a compound interference state. Within a preset observation window after the composite interference state is resolved, the third threshold and the fourth threshold are reduced to preset proportions of their respective nominal values ​​to improve the detection sensitivity of the early degradation state that was obscured during the composite interference state. In response to the absence of a terminal in the early deterioration state among the terminals within the preset observation window, the third threshold and the fourth threshold are restored to their respective nominal values.

[0023] By employing the above technical solution, a combined interference state is identified where the load transition period marker and harmonic distortion rate exceeding the standard simultaneously occur within the same inspection cycle. Under this combined interference state, the load immunity mechanism suspends rate determination, and the harmonic immunity mechanism smooths error data; their simultaneous activation minimizes the system's sensitivity to early degradation signals. Within a limited observation window after the combined interference is resolved, the third and fourth thresholds are reduced to preset proportions of their nominal values, allowing slight degradation signals that might have been masked during the combined interference to be detected during the recovery period. If no abnormality is detected after the observation window ends, the nominal thresholds are automatically restored, preventing prolonged operation at low thresholds from increasing the false alarm rate.

[0024] Optionally, in the threshold mapping relationship, the larger the load current corresponding to the current operating condition level, the larger the second threshold and the smaller the first threshold; The second threshold is determined as follows: the expected temperature rise of each terminal under the current operating condition is calculated based on the load current corresponding to the current operating condition level and the standard contact resistance of each terminal; the expected temperature rise is added to a preset allowable deviation to obtain the second threshold.

[0025] By employing the above technical solution, the expected temperature rise is calculated using the Ohmic heating effect generated by current flow at the terminals under standard contact resistance. A second threshold is then obtained by superimposing an allowable deviation on this value. Under high current conditions, the normal temperature rise of the terminals is higher, resulting in a larger expected temperature rise. Consequently, the second threshold is raised to avoid triggering false alarms due to normal temperature rise. Simultaneously, the signal-to-noise ratio of the self-monitoring signal improves under high current conditions, allowing the first threshold to be lowered to leverage this advantage and enhance error detection sensitivity. Conversely, under low current conditions, the normal temperature rise of the terminals is lower, requiring the second threshold to be lowered to detect slight abnormal temperature rises. This threshold relationship allows the first and second thresholds to adjust inversely with the load current, maintaining a balanced sensitivity for joint judgment across the entire load range.

[0026] In summary, this application includes at least one of the following beneficial technical effects: 1. By performing correlation analysis between self-monitoring error and terminal temperature under the same load condition, the abnormality of the two types of parameters must occur synchronously in time and exceed the first threshold and the second threshold that are dynamically adjusted according to the operating condition level to constitute the abnormality judgment condition. This ensures that unilateral parameter fluctuations caused by non-fault factors will not trigger false alarms. At the same time, the automatic location of the root cause of the fault is achieved by combining the four parameters of dual-channel error and terminal temperature, integrating abnormality detection and fault diagnosis into a closed process.

[0027] 2. By combining the detection of terminal temperature rise rate and inter-terminal temperature difference, early deterioration of terminals can be identified before the absolute value of terminal temperature exceeds the standard. The fault detection window is moved from the post-alarm stage to the early stage when the contact resistance begins to deteriorate. At the same time, the operating condition interference correction mechanism applies corrections for load change, ambient temperature common mode interference and power grid harmonic interference, so that the early deterioration judgment and joint judgment maintain the accuracy of judgment under complex operating conditions.

[0028] 3. By caching temperature data during the load transition period and comparing the expected cumulative temperature rise with the actual cumulative temperature rise after the transition period, by adaptively adjusting the sliding window width according to the distortion rate and discarding extreme interference data under harmonic interference, and by reducing the early degradation judgment threshold within a limited window after the composite interference is removed, the monitoring blind spots generated during the activation of each immune mechanism are compensated, so that the operating condition interference correction can suppress false alarms without missing the real degradation signals that are temporarily blocked. Attached Figure Description

[0029] Figure 1 A flowchart illustrating a power circuit inspection method based on a smart energy meter is shown in one embodiment of the present invention. Detailed Implementation

[0030] The present application will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the application and are not intended to limit the scope of the application.

[0031] This application discloses a power circuit inspection method based on a smart energy meter. Before describing the embodiments of this application in detail, the application scenario and overall technical architecture of this application will be explained first.

[0032] The power circuit inspection method provided in this application is applied to smart meters. A smart meter is an energy metering device that integrates a metering chip, microcontroller, communication module, and temperature sensor, used to meter and monitor the electrical load in a power system. The power circuit of a smart meter refers to the complete signal path from the grid input terminal through sampling devices such as current transformers, current sampling resistors, and voltage divider resistors to the analog-to-digital conversion interface of the metering chip. The health of the power circuit directly affects metering accuracy and electrical safety.

[0033] The terminal block of a smart energy meter has multiple wiring terminals, including current terminals and voltage terminals. The current terminals are used to connect to the input and output lines of a current transformer or current sampling circuit, while the voltage terminals are used to connect to the signal lines of a voltage divider circuit. A temperature sensor is attached to the metal body of each terminal to collect the temperature data of that terminal.

[0034] The inspection method in this application is performed with a fixed inspection cycle. Within each inspection cycle, the smart meter completes one data acquisition and judgment. The time interval of the inspection cycle can be configured according to actual operating requirements, for example, set to 5 seconds.

[0035] As an example, a single-phase smart energy meter with a rated current of 5A will be used for illustration. This energy meter is connected to an industrial power circuit, and its terminal block includes three monitoring points: a current input terminal, a current output terminal, and a voltage terminal. The standard contact resistance of the current input and current output terminals is 0.5mΩ, and the standard contact resistance of the voltage terminal is 0.2mΩ. During a certain inspection cycle, the effective value of the current load current of this energy meter is 3.2A.

[0036] The overall data flow of the inspection method in this application is as follows. First, the effective value of the current load current of the smart energy meter is obtained. The current operating condition level is determined based on the ratio of the effective value of the current load current to the rated current. Then, based on the current operating condition level, a first threshold for judging abnormal self-monitoring error and a second threshold for judging abnormal terminal temperature are determined from a preset threshold mapping relationship. Subsequently, the changes in self-monitoring error of the voltage loop and the current loop of the metering circuit, as well as the changes in terminal temperature of each terminal, are obtained. After obtaining the above data, operating condition interference correction is performed on the changes in self-monitoring error of the voltage loop, the changes in self-monitoring error of the current loop, and the changes in terminal temperature of each terminal to obtain three sets of corrected changes. At the same time, trend analysis is performed on the changes in terminal temperature of each terminal to identify early deterioration terminals. Further, in response to at least one of the corrected error changes exceeding the first threshold and at least one of the corrected terminal temperature changes exceeding the second threshold, or in response to the trend analysis determining that there are terminals in an early deterioration state, a circuit anomaly is determined to exist in the power circuit. Finally, in response to the determination of a circuit anomaly, the fault type is determined based on the combination of the corrected dual-channel error change and the temperature change of each terminal.

[0037] By correlating and analyzing two types of monitoring data—self-monitoring error and terminal temperature—under the same load condition, the abnormality of the two types of parameters must occur synchronously in time to constitute the judgment condition of circuit abnormality. Unilateral parameter fluctuations caused by non-fault factors will not trigger false alarms.

[0038] First step, refer to Figure 1 At the beginning of each inspection cycle, the current effective value of the load current of the smart energy meter is obtained. The current effective value of the load current can be read from the current effective value register of the metering chip, which is automatically updated by the metering chip at the end of each power frequency cycle.

[0039] After obtaining the effective value of the current load current, the ratio of the effective value of the current load current to the rated current of the smart energy meter is used as the load factor. The current operating condition level is determined based on the magnitude of the load factor. For example, if the rated current is 5A, the operating condition level can be divided into four levels: a load factor not exceeding 30% is the light load level, a load factor greater than 30% but not exceeding 80% is the medium load level, a load factor greater than 80% but not exceeding 120% is the heavy load level, and a load factor greater than 120% is the overload level. Taking the above energy meter as an example, the current effective value of the load current is 3.2A, and the load factor is 3.2 / 5 = 64%, falling into the medium load level.

[0040] After determining the current operating condition level, a first threshold and a second threshold are determined from a preset threshold mapping relationship based on the current operating condition level. The first threshold is used to subsequently determine whether the change in self-monitoring error is abnormal, and the second threshold is used to subsequently determine whether the change in terminal temperature is abnormal. The threshold mapping relationship is a set of correspondence tables pre-stored in the non-volatile memory of the smart energy meter, recording the values ​​of the first threshold and the second threshold corresponding to each operating condition level.

[0041] In some embodiments, the larger the load current corresponding to the current operating condition level in the threshold mapping relationship, the larger the second threshold and the smaller the first threshold.

[0042] The second threshold is determined as follows: Based on the load current corresponding to the current operating condition level and the standard contact resistance of each terminal, the expected temperature rise of each terminal under the current operating condition level is calculated. The standard contact resistance is the nominal contact resistance value of the terminal under the factory standard torque, which is written into the non-volatile memory during the energy meter's factory calibration stage. The expected temperature rise is calculated based on the Ohmic thermal effect, that is, the temperature rise caused by Joule heat generated by the contact resistance at the current current level. The expected temperature rise is added to a preset allowable deviation to obtain the second threshold. The allowable deviation is used to cover the measurement error of the temperature sensor and short-term ambient temperature fluctuations.

[0043] Let's take the aforementioned electricity meter as an example. Under medium load conditions, the load current is 3.2A, and the standard contact resistance of the current input terminal is 0.5mΩ. The heating power of this terminal is 3.2² × 0.5 × 10⁻³ = 5.12mW. Based on the thermal resistance parameters of the terminal block, the estimated expected temperature rise is 2.5℃. With an allowable deviation of 1.0℃, the second threshold temperature for the current terminal under medium load conditions is 2.5 + 1.0 = 3.5℃. The standard contact resistance of the voltage terminal is 0.2mΩ. The current flowing through the voltage terminal is much smaller than that of the current terminal, and its expected temperature rise is close to 0. The second threshold temperature is mainly determined by the allowable deviation, for example, set to 1.2℃.

[0044] The first and second thresholds are adjusted inversely according to the operating condition level. Under high current conditions, the normal temperature rise of the terminals is relatively high, and the second threshold needs to be raised accordingly to avoid false alarms triggered by normal temperature rise. At the same time, under high current conditions, the ratio of the amplitude of the self-monitoring injected signal to the amplitude of the sampled signal is relatively stable, and the signal-to-noise ratio of the self-monitoring signal is good. Therefore, the first threshold can be lowered to take advantage of this condition and improve the sensitivity of error detection. Conversely, under low current conditions, the normal temperature rise of the terminals is relatively low, and the second threshold is lowered to capture slight abnormal temperature rises. However, the signal-to-noise ratio of the self-monitoring signal is poor, and the first threshold needs to be raised to avoid false alarms triggered by noise.

[0045] For example, the thresholds corresponding to the four operating condition levels are as follows: under the light load level, the first threshold is 3.0% and the second threshold is 1.5℃; under the medium load level, the first threshold is 2.0% and the second threshold is 3.5℃; under the heavy load level, the first threshold is 1.5% and the second threshold is 5.0℃; under the overload level, the first threshold is 1.2% and the second threshold is 7.0℃.

[0046] In other embodiments, the desired temperature rise can also be obtained through a pre-calibrated temperature rise-current curve, rather than being calculated in real time during each inspection cycle. The calibration process is completed during the meter's factory testing phase: the steady-state temperature rise after the terminals reach thermal equilibrium is recorded at different load current levels, and the correspondence between load current and steady-state temperature rise is written into non-volatile memory in the form of a lookup table. During operation, the desired temperature rise can be obtained directly by looking up the table based on the current load current. Compared to real-time calculation, the lookup table method does not rely on the accuracy of standard contact resistance and thermal resistance parameters, but it requires a complete thermal calibration process to be performed at the factory. Both methods can achieve the function of determining a second threshold based on the desired temperature rise.

[0047] The second step is to obtain the change in self-monitoring error of the voltage loop and the change in self-monitoring error of the current loop of the smart energy meter.

[0048] The metering chip has a built-in self-monitoring reference source, which injects self-monitoring signals into both the current and voltage sampling circuits. The self-monitoring signal is an interharmonic signal with an amplitude of approximately 1% of the rated value and a frequency of several hundred to several kilohertz. After being combined with the normal load current or voltage signal, the self-monitoring signal passes through an external anti-aliasing circuit and enters the metering chip's analog-to-digital converter interface. The metering chip internally extracts the amplitude and phase information of the self-monitoring signal and compares it with the reference value set at the factory calibration to calculate the self-monitoring error. If the device parameters in the sampling circuit remain unchanged, the extracted self-monitoring signal matches the reference value, and the self-monitoring error is zero or close to zero. However, if the parameters of devices such as the current transformer, voltage divider resistor, and anti-aliasing network in the sampling circuit drift, or if the terminal contact resistance changes, the transfer function of the self-monitoring signal changes, and the self-monitoring error increases accordingly.

[0049] The changes in self-monitoring error in the voltage loop and the current loop are two independent monitoring results, corresponding to the voltage sampling channel and the current sampling channel, respectively. Independent acquisition of these two channels is a prerequisite for subsequent fault type determination: the change in self-monitoring error in the current loop is only related to the device parameters and current terminal contact resistance in the current sampling loop, and the change in self-monitoring error in the voltage loop is only related to the device parameters and voltage terminal contact resistance in the voltage sampling loop. If only the synthesized power self-monitoring error is obtained and the voltage and current channels cannot be distinguished, then it is impossible to distinguish between current terminal faults and voltage terminal faults.

[0050] Taking the aforementioned electricity meter as an example, within a certain inspection cycle, the change in self-monitoring error of the voltage loop was 0.3%, and the change in self-monitoring error of the current loop was 0.2%. Both changes are far less than the first threshold of 2.0% under medium load conditions, and are within the normal range of measurement fluctuations.

[0051] The third step is to obtain the temperature change of each terminal of the smart energy meter, including current terminals and voltage terminals.

[0052] The smart meter's terminal block has multiple wiring terminals, and a temperature sensor is attached to the metal body of each terminal. The temperature sensor collects the terminal temperature once per inspection cycle, and the result is the terminal temperature value, expressed in degrees Celsius. The terminal temperature change is the difference between the terminal temperature value of the current inspection cycle and the terminal temperature value of the previous inspection cycle, reflecting the magnitude and direction of the temperature change of that terminal within one inspection cycle.

[0053] Continuing with the example of the aforementioned electricity meter, during a certain inspection cycle, the temperature of the current input terminal was 42.5℃, and the temperature of the previous inspection cycle was 41.7℃. Therefore, the temperature change of the current input terminal is 42.5 - 41.7 = +0.8℃. The temperature of the current output terminal was 42.3℃, and the temperature of the previous inspection cycle was 41.6℃, resulting in a temperature change of +0.7℃. The temperature of the voltage terminal was 38.1℃, and the temperature of the previous inspection cycle was 37.8℃, resulting in a temperature change of +0.3℃.

[0054] Each terminal is labeled as a current terminal and a voltage terminal, respectively. This distinction is the basis for subsequent fault type determination: after determining that the circuit is abnormal, it is necessary to check whether the temperature change of the current terminal and the temperature change of the voltage terminal exceed the second threshold to distinguish whether the fault occurs in the current circuit or the voltage circuit.

[0055] The fourth step involves obtaining the changes in the self-monitoring error of the voltage loop, the self-monitoring error of the current loop, and the terminal temperature changes of each terminal. Then, operating condition interference correction is performed on the above three sets of changes to obtain the corrected changes in the self-monitoring error of the voltage loop, the corrected changes in the self-monitoring error of the current loop, and the corrected terminal temperature changes of each terminal. Finally, trend analysis is performed on the terminal temperature changes of each terminal.

[0056] The purpose of operating condition interference correction is to separate fault signal components from the raw data and suppress operating condition interference components. In actual operating environments, changes in self-monitoring error and terminal temperature are affected not only by loop faults but also by various operating condition factors. For example, sudden changes in load current can cause all terminal temperatures to rise and fall synchronously and rapidly due to the Ohmic heating effect; overall fluctuations in ambient temperature can cause all terminal temperatures to shift synchronously; and harmonics generated by nonlinear loads in the power grid can cause spectral interference near the frequency of the self-monitoring injected signal. If uncorrected raw data is used directly for joint judgment, normal fluctuations caused by operating condition factors may be misjudged as loop faults.

[0057] The specific correction methods included in operating condition interference correction can be selected based on the complexity of the actual operating environment. When the operating environment is relatively stable, such as with gentle load changes, low grid harmonic content, and the electricity meter installed in a temperature-controlled distribution cabinet, correction can be simplified to baseline drift removal of the original data. When the operating environment is complex, such as with frequent load fluctuations, high grid harmonic content, and drastic ambient temperature changes, more comprehensive correction methods are implemented to address different types of operating condition interference.

[0058] Trend analysis and operational disturbance correction are performed in parallel. Trend analysis focuses on the temperature changes of each terminal, identifying terminals in the early deterioration stage by analyzing the temperature change trend over time and the differences between terminals in space. Early deterioration refers to the intermediate state where the terminal contact resistance has begun to increase, but the absolute change in terminal temperature has not yet reached the second threshold. The output of trend analysis serves as the second trigger condition for joint judgment: even if the corrected error and temperature change do not exceed their respective thresholds, the circuit is still considered abnormal as long as the trend analysis determines the presence of an early deteriorated terminal.

[0059] The three sets of corrected changes are then entered into the joint judgment stage, along with the results of the trend analysis. These two sets support the two triggering paths for the joint judgment, which will be further explained in subsequent steps.

[0060] In some embodiments, trend analysis includes the following two sub-steps.

[0061] First, calculate the terminal temperature rise rate between two adjacent inspection cycles. The terminal temperature rise rate is the rate at which the terminal temperature changes over time, reflecting the deteriorating trend of the terminal contact resistance. For terminals operating normally, the temperature rises and falls slowly with changes in load current, and the temperature rise rate remains at a low level. For terminals where contact resistance begins to increase, the temperature rise rate gradually deviates from the normal range due to the continuous accumulation of additional Joule heat.

[0062] Subsequently, the terminal temperature difference between each terminal and the other terminals is calculated. The terminal temperature difference reflects the degree of temperature deviation of that terminal relative to the other terminals within the same meter. All terminals within the same meter share the same ambient temperature. Under stable load conditions, the temperature difference between normal terminals is mainly determined by the differences in standard contact resistance and heat dissipation conditions at the installation location; these values ​​are small and stable. Terminals where contact resistance begins to increase will gradually deviate from the other terminals due to additional Joule heat, resulting in a larger terminal temperature difference.

[0063] A terminal is identified as being in an early degradation state if its temperature rise rate exceeds a third threshold and the temperature difference between that terminal and the other terminals exceeds a fourth threshold. Both conditions must be met simultaneously to trigger early degradation confirmation. A temperature rise rate exceeding the third threshold with a normal temperature difference may indicate a momentary rate jump caused by sampling noise from the temperature sensor, and does not necessarily mean the terminal itself is deteriorating. A temperature difference exceeding the fourth threshold with a normal temperature rise rate may indicate differences in heat dissipation due to different terminal installation positions, or a localized higher temperature due to a terminal being closer to a heat-generating device. Consistent temperature rise rates across all terminals indicate that the contact resistance has not changed. Only when both the temperature rise rate and temperature difference are abnormal does it point to the specific physical event of deteriorating contact resistance: the additional Joule heat generated by increased contact resistance both accelerates the terminal's temperature rise rate (abnormal rate) and gradually widens the temperature gap between the terminal and the other terminals (abnormal temperature difference).

[0064] Taking the aforementioned electricity meter as an example, two sets of comparative scenarios are constructed for illustration. In the normal scenario, the temperature rise rate of the current input terminal is 0.02℃ / s, the temperature difference between it and the current output terminal is 0.2℃, and the temperature difference between it and the voltage terminal is 0.4℃. Assuming the third threshold is 0.05℃ / s and the fourth threshold is 1.5℃, neither condition is met, and the terminal is not in an early deterioration state. In the abnormal scenario, the temperature rise rate of the current input terminal is 0.08℃ / s, exceeding the third threshold of 0.05℃ / s, and the temperature difference between it and the current output terminal is 2.1℃, exceeding the fourth threshold of 1.5℃. Both conditions are met simultaneously, determining that the current input terminal is in an early deterioration state. At this time, the terminal temperature change may not yet exceed the second threshold of 3.5℃, but the deterioration trend of its contact resistance has been detected. The second trigger path, determined by joint judgment, identifies a circuit anomaly, shifting the fault detection window from a post-event alarm after the absolute value of the terminal temperature exceeds the limit to an early stage when the contact resistance begins to deteriorate.

[0065] The above describes one implementation of trend analysis, namely, a two-dimensional joint determination of temperature rise rate and inter-terminal temperature difference. In other embodiments, trend analysis can also be implemented using statistical deviation detection. For each terminal, the historical mean and standard deviation of the terminal temperature change over multiple consecutive inspection cycles are calculated. If the terminal temperature change of a terminal in the current inspection cycle deviates from its historical mean by more than a preset multiple of the standard deviation, the temperature change trend of that terminal is determined to be abnormal. Compared to the joint determination of temperature rise rate and temperature difference, statistical deviation detection does not distinguish between time and spatial dimensions, but uses the historical baseline of each terminal as a reference. This is suitable for scenarios where the number of terminals is small, resulting in limited significance for spatial temperature difference comparisons. For example, when the electricity meter has only two terminals, the temperature difference between the two terminals only reflects the difference between them, and cannot distinguish whether one terminal is abnormally rising or the other is abnormally falling. Statistical deviation detection, however, establishes an independent baseline for each terminal for comparison, without relying on inter-terminal comparisons. Both the joint determination of temperature rise rate and temperature difference and statistical deviation detection can perform trend analysis on the terminal temperature change of each terminal to identify early-deteriorating terminals.

[0066] Furthermore, the calculation process for the terminal temperature rise rate is as follows: Obtain the terminal temperature acquisition value of each terminal in the Kth inspection cycle and the terminal temperature acquisition value in the (K-1)th inspection cycle, where K is a positive integer greater than 1. Divide the difference between the terminal temperature acquisition value in the Kth inspection cycle and the terminal temperature acquisition value in the (K-1)th inspection cycle by the time interval between two adjacent inspection cycles to obtain the terminal temperature rise rate in the Kth inspection cycle.

[0067] Taking the current input terminal of the aforementioned electricity meter as an example, the time interval of the inspection cycle is 5 seconds. Assume the temperature readings of the current input terminal in five consecutive inspection cycles (e.g., K=1 to K=5) are 41.0℃, 41.5℃, 42.0℃, 42.4℃, and 42.5℃, respectively. The temperature rise rate in the K=2nd cycle is (41.5-41.0) / 5=0.10℃ / s, the temperature rise rate in the K=3rd cycle is (42.0-41.5) / 5=0.10℃ / s, the temperature rise rate in the K=4th cycle is (42.4-42.0) / 5=0.08℃ / s, and the temperature rise rate in the K=5th cycle is (42.5-42.4) / 5=0.02℃ / s.

[0068] Due to sampling noise and quantization errors in the temperature sensor, the temperature rise rate calculated cycle by cycle may exhibit significant random fluctuations. For example, in the example above, the temperature rise rate in the K=5th cycle drops sharply from 0.08℃ / s in the previous cycle to 0.02℃ / s. This fluctuation may not be a true change in the terminal state, but rather due to the limited quantization accuracy of the sensor. To suppress this random fluctuation, a sliding window average is performed on the N terminal temperature rise rates calculated over N consecutive inspection cycles to obtain a smoothed terminal temperature rise rate. The smoothed terminal temperature rise rate is then compared with a third threshold, where N is a positive integer greater than or equal to 2.

[0069] Continuing the example above, let N=3. The arithmetic mean of the temperature rise rate for the most recent three cycles (K=3, K=4, K=5) is: (0.10+0.08+0.02) / 3≈0.067℃ / s. This value exceeds the third threshold of 0.05℃ / s, thus proceeding to the subsequent joint temperature difference determination stage. If the sliding window averaging is not performed and the instantaneous temperature rise rate of 0.02℃ / s for the K=5th cycle is used directly, this value does not exceed the third threshold, causing the true degradation trend to be masked by noise from a single cycle.

[0070] The value of N affects the balance between smoothing effect and response speed. A larger N results in a more stable smoothed rate value, better resisting sensor noise, but a slower response to sudden changes in terminal conditions: if a terminal suddenly becomes loose at a certain moment, causing a sharp increase in contact resistance, a larger N means more inspection cycles are needed for the smoothed rate to exceed the third threshold. A smaller N results in a faster response but weaker noise immunity. The value of N can be selected based on the inspection cycle time interval and the quantization accuracy of the temperature sensor. For example, when the inspection cycle is 5 seconds and the quantization accuracy of the temperature sensor is 0.1℃, the quantization step size of the cycle-by-cycle rate is 0.1 / 5 = 0.02℃ / s. An N value of 3 to 5 can achieve a good balance between smoothing effect and response speed.

[0071] In other embodiments, the temperature rise rate can be smoothed using an exponentially weighted moving average instead of a simple arithmetic average. The exponentially weighted moving average assigns greater weight to recent data and less weight to older data, enabling a faster response to sudden changes in terminal states while maintaining smoothness. Compared to the arithmetic average, the exponentially weighted method does not require storing all rate values ​​for the most recent N periods; it only needs to store the smoothed result of the previous period and the instantaneous rate of the current period for updates, making it suitable for embedded platforms with limited storage resources. Both methods can achieve the function of smoothing the terminal temperature rise rate to suppress random fluctuations.

[0072] In some embodiments, operating condition disturbance correction includes at least one of the following: The first item is load mutation immunity. In response to the change in the effective value of the current load current exceeding a preset proportion of the rated current within a single inspection cycle, the inspection cycle is marked as a load transition period. During the load transition period, the determination that the terminal temperature rise rate exceeds the third threshold is suspended.

[0073] When the load current changes drastically, the ohmic heating effect generated by the change in current through all terminals will change rapidly and synchronously. For example, when the load current suddenly increases from 3.2A to 7.5A, the current flowing through the current terminals increases to more than double, and the terminal heating power increases squarely with the current. The temperature rise rate of all current terminals will rise rapidly and synchronously. This increase in the temperature rise rate is a normal physical response to the load change, not a signal of terminal contact resistance degradation. If the rate judgment is not paused, the temperature rise rate of all current terminals during the load change will exceed the third threshold, triggering a large-scale false alarm of early degradation.

[0074] Taking the aforementioned electricity meter as an example, if the load current suddenly changes from 3.2A to 7.5A, the change is 4.3A, exceeding the preset proportion of 60% (3.0A) of the rated current of 5A. This inspection cycle is marked as the load transition period. During the load transition period, even if the temperature rise rate of the current input terminal reaches 0.15℃ / s due to the increased load, far exceeding the third threshold of 0.05℃ / s, it will not trigger the early degradation judgment. The preset proportion must be set greater than the normal load fluctuation range to avoid frequent false triggers, and less than the minimum fault current change detection to avoid missing real load change events.

[0075] The second step is common-mode temperature cancellation. Before determining if the terminal temperature difference exceeds the fourth threshold, the average value of the terminal temperature change of each terminal is calculated. The average value is then subtracted from the terminal temperature change of each terminal to obtain the differential-mode temperature change. The fourth threshold is then determined based on the differential-mode temperature change instead of the terminal temperature difference.

[0076] All terminals within the same meter share the same ambient temperature. When the overall ambient temperature changes, such as a rise in ambient temperature on a summer afternoon, a change in the ventilation conditions of the distribution cabinet, or an increase in the cabinet temperature due to heat generated by adjacent equipment, the temperature changes of all terminals will shift synchronously. This synchronous shift is a common-mode component and does not reflect the difference in contact resistance between terminals. If the original terminal temperature changes are used directly to calculate the temperature difference between terminals, although the common-mode component will theoretically be canceled out by the difference calculation, the residual of the common-mode component may affect the accuracy of the temperature difference comparison due to the gain error of each sensor and the difference in installation position. By first calculating the average value and subtracting the common-mode component, and then calculating the pairwise differences in the differential-mode space, the differential temperature rise between terminals can be reflected more accurately.

[0077] Taking the aforementioned electricity meter as an example, in the afternoon during summer, as the ambient temperature rises, the temperature changes at the three terminals are: current input terminal +1.8℃, current output terminal +1.7℃, and voltage terminal +1.9℃. The average value is (1.8+1.7+1.9) / 3=1.8℃. The differential mode temperature changes at each terminal are: current input terminal 0℃, current output terminal -0.1℃, and voltage terminal +0.1℃. The pairwise differences between the differential mode values ​​do not exceed 0.2℃, which is far less than the fourth threshold of 1.5℃, and does not trigger a temperature difference anomaly. If the original changes are used directly to calculate the temperature difference between terminals, although the difference is small in this example, in extreme cases where sensor gains are inconsistent, such as when one sensor has a higher gain due to aging, the common mode temperature change may be amplified into a false temperature difference signal.

[0078] The third step is harmonic interference smoothing. In response to the total harmonic distortion rate of the grid voltage collected by the smart energy meter exceeding the harmonic interference threshold, the changes in the self-monitoring error of the voltage loop and the self-monitoring error of the current loop are smoothed by taking the median value through a sliding window before joint judgment is performed.

[0079] The total harmonic distortion (THD) of the power grid reflects the content of non-fundamental frequency components in the power grid. When nonlinear loads such as frequency converters and electric arc furnaces are connected to the power grid, the harmonic content increases. The frequency of the self-monitoring injection signal is an interharmonic of several hundred to several thousand hertz, close to some harmonic frequencies. Increased harmonic content introduces interference during the extraction of the self-monitoring signal, causing jumps in the self-monitoring error unrelated to loop faults. If error data affected by harmonic interference is directly used for joint judgment, the jump value may exceed the first threshold, triggering a false alarm when the terminal temperature change also happens to be high for other reasons.

[0080] Taking the aforementioned electricity meter as an example, after industrial users put frequency converters into operation, the total harmonic distortion rate of the power grid voltage increased from 2% to 8%, exceeding the harmonic interference threshold of 5%. A sliding window method was used to smooth the changes in self-monitoring errors of the voltage loop and the current loop. The median has a natural resistance to outliers; even if the error data for a certain period within the window experiences a large jump due to harmonic interference, as long as the data for most periods within the window are normal, the median can still reflect the true error trend.

[0081] Three correction methods protect different parameter dimensions in the judgment process. Load mutation immunity protection for temperature rise rate judgment prevents synchronous rate increases across all terminals caused by load changes from being misjudged as degradation. Common-mode temperature cancellation protection for temperature difference judgment prevents overall ambient temperature fluctuations from interfering with the comparison of differences between terminals. Harmonic interference smoothing protection for self-monitoring error judgment prevents false alarms caused by error jumps triggered by power grid harmonics in the joint judgment. All three methods can be selected for implementation, either fully or partially, depending on the complexity of the actual operating environment.

[0082] Furthermore, the specific process of performing the fourth threshold determination by substituting the differential temperature change value for the terminal temperature difference value is as follows.

[0083] During each inspection cycle, the terminal temperature data for each terminal is collected. Taking the above-mentioned electricity meter as an example, in the Kth inspection cycle, the temperature of the current input terminal is 44.2℃, the temperature of the current output terminal is 43.5℃, and the temperature of the voltage terminal is 39.8℃. In the (K-1)th inspection cycle, the temperature of the current input terminal is 42.1℃, the temperature of the current output terminal is 41.8℃, and the temperature of the voltage terminal is 37.9℃.

[0084] Calculate the difference between the terminal temperature collected at each terminal and the terminal temperature collected in the previous inspection cycle to obtain the terminal temperature change. The terminal temperature change of the current input terminal is 44.2-42.1=+2.1℃, the terminal temperature change of the current output terminal is 43.5-41.8=+1.7℃, and the terminal temperature change of the voltage terminal is 39.8-37.9=+1.9℃.

[0085] The common-mode temperature change component is obtained by averaging the temperature changes of each terminal. This component represents the temperature changes experienced by all terminals within the same meter, including overall increases and decreases in ambient temperature and thermal responses common to all terminals caused by changes in load current. In the example above, the common-mode temperature change component is (2.1 + 1.7 + 1.9) / 3 = 1.9℃.

[0086] Subtract the common-mode temperature change component from the terminal temperature change of each terminal to obtain the differential-mode temperature change of each terminal. The differential-mode temperature change reflects only the unique temperature behavior of that terminal, distinguishing it from the others. In the example above, the differential-mode temperature change of the current input terminal is 2.1 - 1.9 = 0.2℃, the differential-mode temperature change of the current output terminal is 1.7 - 1.9 = -0.2℃, and the differential-mode temperature change of the voltage terminal is 1.9 - 1.9 = 0℃.

[0087] The differential mode temperature change of each terminal is calculated to be the difference between the differential mode temperature change of each terminal and the differential mode temperature change of the other terminals. If the difference exceeds a fourth threshold, a temperature difference anomaly is determined for that terminal. In the example above, the differential mode difference between the current input terminal and the current output terminal is 0.2 - (-0.2) = 0.4℃, the differential mode difference between the current input terminal and the voltage terminal is 0.2 − 0 = 0.2℃, and the differential mode difference between the current output terminal and the voltage terminal is |−0.2−0| = 0.2℃. The fourth threshold is 1.5℃. Since all pairwise differences do not exceed the fourth threshold, a temperature difference anomaly is not triggered. This example demonstrates a normal scenario where an overall increase in ambient temperature leads to relatively large but not significantly different temperature changes at each terminal.

[0088] The following constructs an abnormal scenario where terminal degradation is more pronounced. Assume the current-cycle temperature changes for the three terminals are: current input terminal +3.6℃, ​​current output terminal +1.1℃, and voltage terminal +1.3℃. The common-mode temperature change component = (3.6 + 1.1 + 1.3) / 3 = 2.0℃. The differential-mode temperature changes for each terminal are: current input terminal 3.6 - 2.0 = +1.6℃, current output terminal 1.1 - 2.0 = -0.9℃, and voltage terminal 1.3 - 2.0 = -0.7℃. Calculate the pairwise differences: the differential-mode difference between the current input and current output terminals is 1.6 - (-0.9) = 2.5℃, the differential-mode difference between the current input and voltage terminals is 1.6 - (-0.7) = 2.3℃, and the differential-mode difference between the current output and voltage terminals is |-0.9 - (-0.7)| = 0.2℃. The fourth threshold is 1.5℃. The differential mode difference between the current input terminal and the current output terminal is 2.5℃, exceeding the fourth threshold. The differential mode difference between the current input terminal and the voltage terminal is 2.3℃, also exceeding the fourth threshold, indicating a temperature abnormality at the current input terminal. The differential mode difference between the current output terminal and the voltage terminal is 0.2℃, which does not exceed the fourth threshold, indicating no temperature abnormality between them.

[0089] In the above abnormal example, the common-mode temperature change component is 2.0℃, encompassing the temperature response shared by all terminals due to the overall increase in ambient temperature and changes in load current. After subtracting the common-mode component, the differential modes of the current output terminal and the voltage terminal are -0.9℃ and -0.7℃, respectively. These values ​​are close and both negative, indicating that the temperature changes of these two terminals are below average but the differences between them are small, falling within the normal range. However, the differential mode of the current input terminal is +1.6℃, significantly deviating from zero and being positive. This indicates that this terminal has a unique additional temperature rise component that cannot be explained by ambient temperature or load changes, pointing to an increase in the contact resistance of the terminal itself. If common-mode cancellation is not performed and the original temperature change is used directly to calculate the difference between terminals, the difference between the current input terminal and the current output terminal is 3.6 - 1.1 = 2.5℃, consistent with the result in the differential mode space. However, in practical applications, when the gains of different sensors differ, the reading changes produced by the same common-mode temperature change on different sensors are not entirely the same, and the residual of the common-mode component will be added to the original difference. By first calculating the average value and then subtracting the common-mode component, and then calculating the pairwise differences in the differential-mode space, the interference caused by inconsistent sensor gain can be reduced.

[0090] Furthermore, the management of the load transition period also includes data caching during the transition period, transition period exit determination, and data backtracking verification after exit.

[0091] During the load transition period, although the determination of terminal temperature rise rate exceeding the third threshold was paused, data acquisition did not stop. Terminal temperature data for each terminal continued to be collected and stored as a transition period temperature sequence. This transition period temperature sequence is a set of terminal temperature data arranged chronologically according to the inspection cycle, used for retrospective analysis after the transition period ends.

[0092] The load transition period exit condition is as follows: the load transition period flag is removed when the change in the effective value of the current load current does not exceed a preset proportion of the rated current for M consecutive inspection cycles, where M is a positive integer greater than 1. The requirement that M is greater than 1 is to avoid prematurely removing the transition period flag if the load only momentarily stabilizes and then fluctuates again. A larger M indicates a more conservative exit judgment, requiring the load to remain stable for a longer time window before the transition period is removed, but the duration of the transition period is also correspondingly extended.

[0093] Taking the aforementioned electricity meter as an example, the load current suddenly jumps from 3.2A to 7.5A, triggering the load transition period marker. Afterward, the load current gradually stabilizes over several inspection cycles: 7.5A in the first cycle, 7.3A in the second, 7.2A in the third, 7.2A in the fourth, 7.1A in the fifth, and 7.2A in the sixth. From the fourth cycle onward, the current changes in adjacent cycles are 0A, -0.1A, and +0.1A, respectively, all not exceeding the preset 60% threshold of 3.0A. Assuming M=3, the changes in current for three consecutive cycles (cycles 4, 5, and 6) from the fourth cycle onward do not exceed the limit, thus releasing the load transition period marker. The transition period lasts for a total of 6 inspection cycles, or 30 seconds.

[0094] After the load transition period marker is removed, data backtracking verification is performed. While the pause rate determination avoids false alarms, it also introduces a monitoring blind spot: if a terminal happens to begin deteriorating during a load surge, its deterioration signal is masked by the pause determination. The purpose of backtracking verification is to compensate for this blind spot.

[0095] Specifically, the expected cumulative temperature rise of each terminal during the load transition period is calculated based on the current operating condition level corresponding to each inspection cycle within the load transition period. The expected cumulative temperature rise refers to the total cumulative temperature rise that a normal terminal with standard contact resistance should experience under load conditions during each inspection cycle within the transition period. Since the load current changes continuously during the transition period, the operating condition level corresponding to different inspection cycles may be different. Therefore, it is necessary to calculate the expected temperature rise for each cycle and then sum them up.

[0096] Continuing with the example above, the transition period consists of 6 cycles. During these 6 cycles, the load current ranges from 7.1A to 7.5A, with load rates exceeding 120%, corresponding to overload levels. The expected temperature rise per cycle is approximately 0.8℃. The expected cumulative temperature rise at the current-input terminals is 6 × 0.8 = 4.8℃.

[0097] The actual cumulative temperature rise of each terminal in the transition temperature sequence is compared with the expected cumulative temperature rise. The actual cumulative temperature rise is calculated from the transition temperature sequence: the temperature acquisition value of the last cycle of the transition period minus the temperature acquisition value of the first cycle of the transition period. Continuing the example above, the temperature acquisition value of the current input terminal in the first cycle of the transition period is 43.0℃, and the temperature acquisition value in the sixth cycle of the transition period is 49.8℃. The actual cumulative temperature rise is 49.8 - 43.0 = 6.8℃.

[0098] If the actual cumulative temperature rise of a terminal exceeds the sum of the expected cumulative temperature rise and a preset deviation, it is determined that the terminal has a superimposed degradation signal during the load transition period. The preset deviation is used to cover model errors and sensor accuracy errors in the expected temperature rise calculation, and is set to, for example, 0.8℃. In the example above, the judgment threshold for the current-input terminal is 4.8 + 0.8 = 5.6℃. The actual cumulative temperature rise of 6.8℃ exceeds the judgment threshold, confirming that the terminal has a superimposed degradation signal during the transition period. The actual cumulative temperature rise of a normal terminal is determined by the Ohmic thermal effect caused by load changes and matches the expected value calculated based on the standard contact resistance. For degraded terminals where the contact resistance has begun to increase, the actual cumulative temperature rise will systematically exceed the expected value; the excess is the superimposed temperature rise component caused by degradation.

[0099] Terminals with superimposed degradation signals are identified as being in an early degradation state, enter the second trigger path of joint determination, and then trigger the fault type determination.

[0100] Furthermore, the specific process of taking the median value of the sliding window in harmonic interference smoothing is as follows: The harmonic interference level is determined based on the magnitude of the total harmonic distortion (THD) of the power grid voltage. The harmonic interference level reflects the intensity of interference from power grid harmonics to the self-monitoring signal extraction. The harmonic interference level can be divided into three levels: a THD of no more than 5% is considered normal (smoothing not enabled); a THD greater than 5% but no more than 10% is considered mild interference; a THD greater than 10% but no more than 20% is considered moderate interference; and a THD greater than 20% is considered extreme interference.

[0101] The width of the sliding window is selected from a preset window width mapping relationship based on the harmonic interference level. The higher the harmonic interference level, the wider the sliding window. A wider window accommodates more inspection cycles, making the median calculation more resistant to individual outliers. However, the timeliness of the data deteriorates because the earliest data included in the window is far removed from the current time, while the median reflects a trend over a longer period rather than the instantaneous state at the current moment. For example, the window width is 3 inspection cycles for a mild interference level and 5 inspection cycles for a moderate interference level.

[0102] Taking the aforementioned electricity meter as an example, the total harmonic distortion rate of the grid voltage increased from 2% to 12%, falling into the moderate interference level. The window width was set to 5 inspection cycles. The changes in voltage loop self-monitoring error in the most recent 5 cycles were 0.4%, 0.5%, 2.8%, 0.6%, and 0.5%, respectively. Among them, the 2.8% in the 3rd cycle was an abnormal jump caused by harmonic interference. Taking the median of these 5 data points: the data were sorted as 0.4%, 0.5%, 0.5%, 0.6%, and 2.8%, with a median of 0.5%. The abnormal jump value of 2.8% was naturally filtered out by the median calculation because it was at the end of the sorted data. The smoothed data of 0.5% can reflect the true error trend. If the mean value is used instead of the median value, the average value is (0.4+0.5+2.8+0.6+0.5) / 5=0.96%, and the abnormal jump value will pull the mean value up to close to 1%, deviating from the true trend.

[0103] In response to the total harmonic distortion (THD) of the grid voltage exceeding the extreme interference threshold, the changes in self-monitoring errors of the voltage loop and current loop collected during the corresponding inspection cycle are marked as invalid data. Invalid data is not included in the median calculation of the sliding window. The extreme interference threshold corresponds to scenarios with extremely high harmonic distortion rates, such as a THD exceeding 20%. Under these conditions, the extraction process of the self-monitoring signal is severely affected by spectral interference, and the error data completely loses its physical meaning; no statistical processing can recover the true error information. If this distorted data is included in the sliding window, even with median calculation, when most data within the window comes from the extreme interference cycle, the median itself is the median of the distorted data and has no reference value.

[0104] Taking the aforementioned electricity meter as an example, let's assume the extreme interference threshold is 20%. If the total harmonic distortion (THD) of the grid voltage reaches 25% within a certain inspection cycle, the changes in the self-monitoring error of the voltage loop and the self-monitoring error of the current loop for that cycle are marked as invalid data. Assuming a window width of 5, if 2 out of 5 cycles within the window are marked as invalid, the median calculation is based solely on the data from the remaining 3 valid cycles. If all 5 cycles within the window are marked as invalid, meaning the THD exceeds 20% for 5 consecutive cycles, then there is no valid data available for calculation within that window. The joint judgment is temporarily suspended in the current cycle, awaiting the accumulation of subsequent valid data before resuming the judgment.

[0105] The tiered response allows the smoothing strategy to automatically adjust the trade-off between timeliness and reliability according to the intensity of the disturbance. Under mild disturbance, the window is narrower to preserve the timeliness of the error data; under moderate disturbance, the window is widened to enhance the smoothing effect; and under extreme disturbance, distorted data is discarded directly to ensure that all data entering the joint decision-making process has physical meaning.

[0106] In some embodiments, the inspection method further includes the identification and coordinated processing of complex interference states.

[0107] If the load transition period flag and the total harmonic distortion rate of the grid voltage exceeding the harmonic interference threshold both occur simultaneously within the same inspection cycle, it is determined that the inspection cycle is in a compound interference state. A compound interference state is a specific operating scenario where both load change immunity and harmonic interference smoothing correction methods are activated simultaneously. In industrial settings, this scenario is not an extremely rare event, but rather a high-frequency actual operating condition: the switching of high-power nonlinear loads can simultaneously trigger sudden changes in load current and a sharp increase in grid harmonic content. For example, during inverter startup, as the motor accelerates from zero speed to rated speed, the current rapidly rises from zero to several times the rated value, while the inverter's rectifier bridge and inverter bridge inject a large amount of harmonic current into the grid. The operation of an electric arc furnace also produces a superposition effect of sudden current changes and harmonic injection.

[0108] Under combined interference conditions, the load transition immunity pauses the determination of terminal temperature rise rate exceeding the third threshold, and harmonic interference smoothing performs median filtering or even discards self-monitoring error data. The simultaneous activation of both mechanisms means that the system's sensitivity to early degradation signals is at its lowest level: rate-dimensional detection is paused, and error-dimensional data is smoothed or discarded. If a terminal happens to begin deteriorating during combined interference, the degradation signal is masked in both dimensions, resulting in the lowest detection probability.

[0109] To compensate for the loss of detection sensitivity during the combined interference, within a preset observation window after the combined interference is resolved, the third and fourth thresholds are reduced to preset proportions of their respective nominal values. A lower third threshold means greater sensitivity in determining the temperature rise rate, enabling the detection of more subtle rate anomalies; a lower fourth threshold means greater sensitivity in determining temperature differences, enabling the detection of more subtle temperature deviations between terminals. This limited-time sensitivity enhancement compensation allows subtle degradation signals that might be masked during the combined interference to be detected during the recovery period after the interference subsides.

[0110] The above-mentioned electricity meter serves as an example. When the industrial frequency converter is put into operation, the load current jumps from 3.2A to 8.0A. This change of 4.8A exceeds the preset ratio of 3.0A, triggering the load transition period flag. Simultaneously, the total harmonic distortion rate of the power grid voltage increases from 2% to 15%, exceeding the harmonic interference threshold of 5%. Both conditions are met within the same inspection cycle, marking that cycle as a composite interference state. After several cycles, the frequency converter reaches steady-state operation, and the load current stabilizes around 7.5A, achieving the condition that the change has not exceeded the standard for M consecutive cycles. The load transition period flag is then lifted, and the power grid harmonic content drops back to 6%, which is still considered higher than normal but lower than the extreme interference threshold. Therefore, the composite interference state is lifted.

[0111] After the composite interference is resolved, the system enters a preset observation window. This window is set to 10 inspection cycles (50 seconds). Within this window, the third threshold decreases from the nominal value of 0.05℃ / s to 70% of the nominal value, i.e., 0.035℃ / s. The fourth threshold decreases from the nominal value of 1.5℃ to 70% of the nominal value, i.e., 1.05℃. If the current-input terminal does indeed begin to deteriorate during the composite interference period, its temperature rise rate may only be 0.04℃ / s, and its temperature difference may only be 1.2℃. Here, we determine that the temperature rise rate does not exceed the nominal third threshold of 0.05℃ / s, but exceeds the reduced 0.035℃ / s; the temperature difference does not exceed the nominal fourth threshold of 1.5℃, but exceeds the reduced 1.05℃. Therefore, since both conditions are met simultaneously within the sensitivity enhancement window, the terminal is confirmed to be in an early deterioration state.

[0112] In response to the absence of any terminals in an early deterioration state within the preset observation window, the third and fourth thresholds are restored to their respective nominal values. The automatic recovery mechanism is a key constraint for controlling false alarm risk—lowering the threshold increases sensitivity, but also increases the possibility of misinterpreting normal fluctuations as deterioration. The sensitivity enhancement timeframe is strictly limited to the preset observation window after the removal of composite interference, ensuring that sensitivity enhancement only takes effect during the period most likely to experience masking degradation. After the window ends, the nominal thresholds are restored to maintain the false alarm control level expected during normal operation.

[0113] The length of the preset observation window and the threshold reduction ratio need to strike a balance between detection sensitivity and the risk of false alarms. The longer the window and the larger the reduction ratio, the wider the coverage of the compensated detection and the higher the sensitivity, but the higher the probability of misjudging normal fluctuations as degradation within the window period. The shorter the window and the smaller the reduction ratio, the lower the risk of false alarms, but it may miss degradation signals that have begun but are progressing slowly during the period of compound interference.

[0114] Fifth, after completing the operating condition interference correction and trend analysis, perform a joint judgment: The joint determination includes two triggering paths; satisfying either one determines that there is a circuit anomaly in the power circuit.

[0115] The first path is a dual-parameter joint over-limit trigger. A loop anomaly is determined when at least one of the corrected voltage loop self-monitoring error change and the corrected current loop self-monitoring error change exceeds a first threshold, and at least one of the corrected terminal temperature changes exceeds a second threshold. This path requires both self-monitoring error and terminal temperature to simultaneously meet their respective threshold conditions within the same inspection cycle. An excessive self-monitoring error indicates a change in the signal transfer function of the sampling loop, while an excessive terminal temperature indicates a change in the heat generation of a physical node in the loop. The synchronous anomaly of both parameters points to the same physical event—a fault in a certain part of the loop that causes simultaneous changes in electrical and thermal parameters. If only the self-monitoring error exceeds the limit while the terminal temperature is normal, it may be due to load spectrum interference or internal fluctuations in the metering chip causing error jumps, and does not constitute a loop fault. If only the terminal temperature exceeds the limit while the self-monitoring error is normal, it may be due to increased ambient temperature or temperature shifts caused by radiant heat from adjacent equipment, and similarly does not constitute a loop fault. Cross-validation of the two types of parameters ensures that unilateral parameter fluctuations caused by non-fault factors will not trigger false alarms.

[0116] The design intent of "at least one" in the first path is that the voltage loop and the current loop are two independent sampling channels. An abnormality in either channel indicates that the overall health of the metering loop has been affected. For example, an abnormal contact at the current terminal may only cause the change in the self-monitoring error of the current loop to exceed the standard, while the change in the self-monitoring error of the voltage loop remains normal, but the entire loop already has a fault. If both channels are required to exceed the standard simultaneously before triggering, a single-channel fault will be missed.

[0117] Taking the aforementioned electricity meter as an example, three scenarios will be constructed for illustration.

[0118] Scenario 1: This scenario corresponds to the case where neither of the two trigger paths is triggered. The change in the self-monitoring error of the current loop after correction is 1.5%, which does not exceed the first threshold of 2.0% under medium load conditions; the change in the temperature of the current input terminal after correction is 2.8℃, which does not exceed the second threshold of 3.5℃; the trend analysis did not find any early deterioration terminals. Neither path meets the requirements, so the loop abnormality is not triggered, and the next inspection cycle continues.

[0119] Scenario 2: This scenario corresponds to the triggering of the first path. The self-monitoring error change of the corrected current loop is 3.2%, exceeding the first threshold of 2.0%; the temperature change of the corrected current input terminal is 4.1℃, exceeding the second threshold of 3.5℃. With both error and temperature parameters exceeding the limits simultaneously, the loop is determined to be abnormal, and the fault type determination process begins.

[0120] Scenario 3: This scenario corresponds to the triggering of the second path. After correction, the change in the self-monitoring error of the current loop is 1.2%, which does not exceed the first threshold; the temperature change of each terminal after correction does not exceed the second threshold. However, trend analysis determines that the current input terminal is in an early deterioration state, i.e., the temperature rise rate exceeds the third threshold and the temperature difference exceeds the fourth threshold. Although the first path is not met, the second path is, thus determining that the loop is abnormal and entering the fault type determination stage.

[0121] The second path is early degradation triggering. In response to trend analysis identifying terminals in an early degradation state, a circuit anomaly is determined. This path targets the stage where terminal contact resistance has begun to deteriorate but has not yet reached the point where both self-monitoring error and terminal temperature absolute value exceed limits simultaneously. At this stage, the terminal temperature change trend begins to show an accelerated rate and widening temperature difference, but the absolute value of the change is not yet sufficient to trigger the first path. The second path shifts the fault detection window from a reactive alarm after the absolute value exceeds the limit to an early stage of abnormal trend.

[0122] The two paths are linked by an "OR" relationship, covering different stages from the initial stage to the development of a fault. The early stage is covered by the second path, and the development stage is covered by the first path.

[0123] In response to the determination that there is a circuit anomaly in the power circuit, the fault type determination process begins.

[0124] The sixth step involves determining the fault type based on the combination of the corrected voltage loop self-monitoring error change, the corrected current loop self-monitoring error change, and the corrected terminal temperature changes of the current and voltage terminals. Fault type determination uses four input parameters: the corrected error changes of two independent channels and the corrected temperature changes of the two types of terminals. The combination of these four parameters reflects the specific location and nature of the fault.

[0125] The fault type determination logic is based on the physical causal relationship between the self-monitoring error channel attribution and the terminal temperature response. When a fault occurs in a certain link of the circuit, the response patterns of the self-monitoring error and the terminal temperature depend on the physical location and nature of the fault: an increase in terminal contact resistance changes the signal transfer function of the corresponding sampling circuit, leading to an increase in error, and also generates additional Joule heat on the contact surface, leading to a temperature increase; both have the characteristic of co-occurring on the same side; drift of internal parameters of the sampling device only changes the signal transfer function and does not generate additional heat on the terminal contact surface; drift of the metering chip reference simultaneously affects both self-monitoring reference sources, but does not involve any physical degradation of external circuit devices.

[0126] In some embodiments, fault type determination can be achieved by looking up a preset fault feature mapping table. The mapping table uses the state combinations of four parameters as indexes and fault type labels as outputs. The mapping table is written into non-volatile memory when the energy meter leaves the factory. In other embodiments, fault type determination is implemented through conditional branching logic, checking each state combination of the four parameters and matching it with the corresponding fault type. Both methods can achieve the function of determining the fault type based on the combination of four parameters.

[0127] In some embodiments, the fault type is determined through the following five conditional branches: The first branch, responding to the fact that the change in the corrected current loop self-monitoring error exceeds a first threshold and the change in the temperature of the current terminal among the corrected terminal temperature changes exceeds a second threshold, determines the fault type as abnormal current terminal contact. Increased contact resistance of the current terminal produces two effects simultaneously: firstly, the increased contact resistance alters the equivalent impedance of the current sampling loop, changing the transfer function of the self-monitoring signal as it passes through the loop, leading to an increase in the change in the current loop self-monitoring error; secondly, increased contact resistance under current-carrying conditions increases Joule heating on the contact surface, causing the current terminal temperature to rise. Therefore, the simultaneous occurrence of excessive current loop error and excessive current terminal temperature on the same side points to abnormal current terminal contact.

[0128] Taking the aforementioned electricity meter as an example, the contact surface of the current input terminal oxidizes due to long-term operation, increasing the contact resistance from the standard value of 0.5mΩ to 2.0mΩ. Under medium-load conditions, the load current is 3.2A, and the heat generation power of this terminal increases from 3.2²×0.5×10⁻³=5.12mW to 3.2²×2.0×10⁻³=20.48mW, a significant increase in temperature rise. The corrected current loop self-monitoring error change is 3.2%, exceeding the first threshold of 2.0%; the corrected current input terminal temperature change is 4.1℃, exceeding the second threshold of 3.5℃. The corrected voltage loop self-monitoring error change is 0.4%, not exceeding the first threshold; the corrected voltage terminal temperature change is 0.5℃, not exceeding the second threshold. The fault type is determined to be abnormal contact of the current terminal.

[0129] The second branch, in response to the self-monitoring error change of the corrected voltage loop exceeding a first threshold and the terminal temperature change of the voltage terminal exceeding a second threshold among the terminal temperature changes of each corrected terminal, determines the fault type as abnormal voltage terminal contact. The physical mechanism is the same as the first branch, the difference being that the fault occurs on the voltage sampling circuit side. Increased contact resistance of the voltage terminal causes a change in the voltage division ratio of the voltage divider circuit, thereby affecting the transmission of the self-monitoring signal of the voltage loop. Simultaneously, heating of the contact surface causes the voltage terminal temperature to rise.

[0130] Taking the aforementioned electricity meter as an example, the voltage terminals experienced increased contact resistance due to loose bolts. The corrected voltage loop self-monitoring error change was 2.5%, exceeding the first threshold of 2.0%; the corrected voltage terminal temperature change was 1.8℃, exceeding the second threshold of 1.2℃ under light load conditions. The corrected current loop self-monitoring error change was 0.3%, not exceeding the first threshold; the corrected current terminal temperature change was 0.6℃, not exceeding the second threshold. The fault type was determined to be abnormal voltage terminal contact.

[0131] The third branch determines the fault type as current sampling device degradation when the corrected change in the current loop self-monitoring error exceeds the first threshold and the corrected terminal temperature change of the current terminals does not exceed the second threshold. The current sampling device includes a current transformer and RC components in the anti-aliasing network. When a short circuit occurs between winding turns in the current transformer, or the core magnetic characteristics degrade, or the resistance value in the anti-aliasing network drifts, the signal transfer function of the current sampling loop changes, leading to an increase in the change in the current loop self-monitoring error. However, internal faults in these devices do not generate additional Joule heat at the terminal contact surfaces—the heat generated by the short circuit between winding turns is absorbed by the transformer's own heat dissipation structure and is not conducted to the terminal block. Therefore, the combination of excessive current loop error but normal current terminal temperature points to current sampling device degradation.

[0132] Taking the aforementioned electricity meter as an example, the current transformer's core permeability decreased due to long-term operation, causing the mutual inductance ratio to deviate from the nominal value. The corrected self-monitoring error change in the current loop was 2.8%, exceeding the first threshold of 2.0%; the corrected temperature change at the current input terminal was 0.9℃, not exceeding the second threshold of 3.5℃; the corrected temperature change at the current output terminal was 0.8℃, not exceeding the second threshold. The corrected self-monitoring error change in the voltage loop was 0.3%, not exceeding the first threshold. The fault type was determined to be deterioration of the current sampling device. Maintenance personnel can specifically inspect the current transformer and the resistive and capacitive components in the current sampling loop without needing to check the terminal contact surfaces.

[0133] The fourth branch determines the fault type as voltage sampling device degradation when the corrected voltage loop self-monitoring error exceeds the first threshold and the corrected terminal temperature change of the voltage terminals does not exceed the second threshold. The physical mechanism is the same as the third branch, except the fault occurs on the voltage sampling loop side. Resistance drift of the voltage divider resistor or changes in the parameters of the anti-aliasing network increase the voltage loop self-monitoring error but do not generate additional heat at the voltage terminal contact surfaces.

[0134] Taking the aforementioned electricity meter as an example, the voltage divider resistor experienced resistance deviation due to temperature cycling. The corrected voltage loop self-monitoring error change was 2.3%, exceeding the first threshold of 2.0%; the corrected voltage terminal temperature change was 0.4℃, not exceeding the second threshold of 1.2℃. The corrected current loop self-monitoring error change was 0.2%, not exceeding the first threshold. The fault type was determined to be voltage sampling device degradation.

[0135] The fifth branch, responding to the fact that both the corrected current loop self-monitoring error change and the corrected voltage loop self-monitoring error change exceed the first threshold, while the corrected terminal temperature changes do not exceed the second threshold, determines the fault type as metering chip reference drift. The self-monitoring reference source inside the metering chip provides a unified reference for the self-monitoring signal injection of the voltage and current loops. When the output voltage of the reference source shifts due to chip aging, temperature drift, or other reasons, the amplitude of the self-monitoring signal injected into the two sampling loops changes simultaneously, causing the self-monitoring error changes of both channels to increase simultaneously. However, reference source drift does not involve any physical degradation of external loop components; the terminal contact surfaces and sampling devices remain unchanged, therefore the temperature changes of all terminals remain normal. The combination of simultaneous exceedance of dual-channel errors while the temperature of all terminals is normal eliminates external loop causes, locking the fault within the metering chip itself.

[0136] Taking the aforementioned electricity meter as an example, the metering chip experienced reference source voltage drift due to long-term operation. The self-monitoring error change in the corrected current loop was 2.8%, exceeding the first threshold of 2.0%; the self-monitoring error change in the corrected voltage loop was 2.5%, also exceeding the first threshold of 2.0%. The temperature changes at the corrected current input terminals were 0.6℃, current output terminals were 0.5℃, and voltage terminals were 0.3℃, all within their respective second thresholds. The fault type was determined to be reference drift in the metering chip. Maintenance personnel can directly replace the metering chip or the entire meter without needing to check the terminals and sampling devices.

[0137] Five branches exhaustively list all possible temperature state combinations under the premise that "at least one error channel exceeds the limit". When the current loop exceeds the limit, the current terminal temperature exceeds the limit, corresponding to the first branch; when the current terminal temperature is normal, corresponding to the third branch. When the voltage loop exceeds the limit, the voltage terminal temperature exceeds the limit, corresponding to the second branch; when the voltage terminal temperature is normal, corresponding to the fourth branch. When both channels exceed the limit simultaneously and all terminal temperatures are normal, corresponding to the fifth branch. The judgment conclusion for each fault type is exclusive, and maintenance personnel can directly locate the faulty link based on the output fault type label.

[0138] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for inspecting power circuits based on smart meters, characterized in that, include: S1. Obtain the current effective value of the load current, determine the current operating condition level based on the ratio of the current effective value of the load current to the rated current, and determine the first threshold and the second threshold from the preset threshold mapping relationship based on the current operating condition level; S2. Obtain the change in self-monitoring error of the voltage loop and the change in self-monitoring error of the current loop in the metering circuit; S3. Obtain the terminal temperature change of each terminal, wherein each terminal includes a current terminal and a voltage terminal; S4. Perform operating condition interference correction on the voltage loop self-monitoring error change, the current loop self-monitoring error change, and the terminal temperature change of each terminal to obtain the corrected voltage loop self-monitoring error change, the corrected current loop self-monitoring error change, and the corrected terminal temperature change of each terminal, and perform trend analysis on the terminal temperature change of each terminal. S5. In response to at least one of the corrected voltage loop self-monitoring error change and the corrected current loop self-monitoring error change exceeding the first threshold and at least one of the corrected terminal temperature changes of each terminal exceeding the second threshold, or in response to determining, based on the trend analysis, that there is a terminal in an early deterioration state among the terminals, it is determined that there is a loop anomaly in the power circuit of the smart energy meter. S6. In response to determining that the power circuit has the circuit abnormality, the fault type is determined based on the combination of the corrected voltage circuit self-monitoring error change, the current circuit self-monitoring error change, and the terminal temperature change of each terminal, specifically the terminal temperature change of the current terminal and the terminal temperature change of the voltage terminal.

2. The power circuit inspection method according to claim 1, characterized in that, S6 includes the following sub-steps: S61. In response to the fact that the change in the self-monitoring error of the corrected current loop exceeds the first threshold and the change in the terminal temperature of the current terminal in the terminal temperature change of each corrected terminal exceeds the second threshold, the fault type is determined to be abnormal contact of the current terminal. S62. In response to the fact that the change in the self-monitoring error of the corrected voltage loop exceeds the first threshold and the change in the terminal temperature of the voltage terminal among the changes in the terminal temperature of each corrected terminal exceeds the second threshold, the fault type is determined to be abnormal voltage terminal contact. S63. In response to the fact that the change in the self-monitoring error of the corrected current loop exceeds the first threshold and the change in the terminal temperature of the current terminal among the changes in the terminal temperature of each corrected terminal does not exceed the second threshold, the fault type is determined to be current sampling device degradation. S64. In response to the fact that the change in the self-monitoring error of the corrected voltage loop exceeds the first threshold and the change in the terminal temperature of the voltage terminal among the changes in the terminal temperature of each corrected terminal does not exceed the second threshold, the fault type is determined to be voltage sampling device degradation. S65. In response to the fact that both the corrected change in the self-monitoring error of the current loop and the corrected change in the self-monitoring error of the voltage loop exceed the first threshold and the corrected change in the terminal temperature of each terminal does not exceed the second threshold, the fault type is determined to be metering chip reference drift.

3. The power circuit inspection method according to claim 2, characterized in that, The trend analysis of the terminal temperature change performed on each terminal in S4 includes: S41. Calculate the terminal temperature rise rate between two adjacent inspection cycles; S42. Calculate the terminal temperature difference between each terminal and the other terminals. In response to the terminal temperature rise rate of a terminal exceeding a third threshold and the terminal temperature difference between the terminal and the other terminals exceeding a fourth threshold, determine that the terminal is in an early deterioration state.

4. The power circuit inspection method according to claim 3, characterized in that, S41 includes the following sub-steps: S411. Obtain the terminal temperature acquisition value of each terminal in the Kth inspection cycle and the terminal temperature acquisition value in the (K-1)th inspection cycle, where K is a positive integer greater than 1; S412. Divide the difference between the terminal temperature acquisition value of the Kth inspection cycle and the terminal temperature acquisition value of the (K-1)th inspection cycle by the time interval between two adjacent inspection cycles to obtain the terminal temperature rise rate. S413. Perform a sliding window average on the N terminal temperature rise rates calculated over N consecutive inspection cycles to obtain a smoothed terminal temperature rise rate, and compare the smoothed terminal temperature rise rate with the third threshold, where N is a positive integer greater than or equal to 2.

5. The power circuit inspection method according to claim 3, characterized in that, The operating condition interference correction in S4 includes at least one of the following: In response to the change in the effective value of the current load current exceeding a preset proportion of the rated current within a single inspection cycle, a mark is added to the inspection cycle as a load transition period, and the determination that the terminal temperature rise rate exceeds the third threshold is suspended during the load transition period. Before determining that the terminal temperature difference exceeds the fourth threshold, the average value of the terminal temperature change of each terminal is calculated, and the differential temperature change is obtained by subtracting the average value from the terminal temperature change of each terminal. The fourth threshold is then determined based on the differential temperature change instead of the terminal temperature difference. In response to the total harmonic distortion rate of the grid voltage collected by the smart energy meter exceeding the harmonic interference threshold, the change in the self-monitoring error of the voltage loop and the change in the self-monitoring error of the current loop are smoothed by taking the median value through a sliding window before the determination in S5 is executed.

6. The power circuit inspection method according to claim 5, characterized in that, The step of performing the determination of the fourth threshold based on the differential mode temperature change instead of the terminal temperature difference value includes the following sub-steps: S71. During each inspection cycle, acquire the terminal temperature data of each terminal; S72. Calculate the difference between the terminal temperature acquisition value of each terminal and the terminal temperature acquisition value of the previous inspection cycle to obtain the terminal temperature change of each terminal. S73. Calculate the average value of the terminal temperature change of each terminal to obtain the common mode temperature change component; S74. Subtract the common-mode temperature change component from the terminal temperature change of each terminal to obtain the differential-mode temperature change of each terminal; S75. Calculate the difference between the differential mode temperature change of each terminal and the differential mode temperature change of the other terminals, and determine the temperature difference abnormality of the terminal in response to the difference exceeding the fourth threshold.

7. The power circuit inspection method according to claim 5, characterized in that, The addition of a mark for a load transition period to the inspection cycle also includes: During the load transition period, the terminal temperature data of each terminal is continuously collected and stored as a transition period temperature sequence. In response to the fact that the change in the effective value of the current load current does not exceed the preset ratio of the rated current for M consecutive inspection cycles, the load transition period marker is removed, where M is a positive integer greater than 1; After removing the mark of the load transition period, the expected cumulative temperature rise value of each terminal during the load transition period is calculated according to the current operating condition level corresponding to each inspection cycle during the load transition period. The actual cumulative temperature rise value of each terminal in the transition period temperature sequence is compared with the expected cumulative temperature rise value. In response to the fact that the actual cumulative temperature rise of a terminal exceeds the sum of the expected cumulative temperature rise and a preset deviation, it is determined that the terminal has a superimposed degradation signal during the load transition period, and the terminal with the superimposed degradation signal is determined to be in the early degradation state.

8. The power circuit inspection method according to claim 5, characterized in that, The smoothing of the self-monitoring error changes of the voltage loop and the self-monitoring error changes of the current loop using a sliding window midpoint method includes: The harmonic interference level is determined based on the magnitude of the total harmonic distortion rate of the power grid voltage. The width of the sliding window is selected from a preset window width mapping relationship according to the harmonic interference level, wherein the higher the harmonic interference level, the larger the width of the sliding window. In response to the total harmonic distortion rate of the grid voltage exceeding the extreme interference threshold, the changes in the self-monitoring error of the voltage loop and the self-monitoring error of the current loop collected in the corresponding inspection cycle are marked as invalid data, and the invalid data is not included in the median calculation of the sliding window.

9. The power circuit inspection method according to claim 5, characterized in that, Also includes: If the load transition period marker and the total harmonic distortion rate of the grid voltage exceeding the harmonic interference threshold are both established within the same inspection cycle, it is determined that the inspection cycle is in a compound interference state. Within a preset observation window after the composite interference state is resolved, the third threshold and the fourth threshold are reduced to preset proportions of their respective nominal values ​​to improve the detection sensitivity of the early degradation state that was obscured during the composite interference state. In response to the absence of a terminal in the early deterioration state among the terminals within the preset observation window, the third threshold and the fourth threshold are restored to their respective nominal values.

10. The power circuit inspection method according to claim 1, characterized in that, In the threshold mapping relationship, the larger the load current corresponding to the current operating condition level, the larger the second threshold and the smaller the first threshold. The second threshold is determined as follows: the expected temperature rise of each terminal under the current operating condition is calculated based on the load current corresponding to the current operating condition level and the standard contact resistance of each terminal; the expected temperature rise is added to a preset allowable deviation to obtain the second threshold.

Images