An automatic inspection system for temperature rise of connection terminals of an electric energy metering box

Abstract

The present application belongs to the technical field of fault detection, and particularly relates to a kind of electric energy metering box connection terminal temperature rise automatic inspection system, including acquisition module, obtains real-time temperature, ambient temperature and load current;Feature extraction module generates thermal state feature vector containing temperature absolute value, temperature rise rate, duration and contact resistance change trend factor;Abnormality discrimination module compares feature vector with adaptively updated dynamic threshold set, and outputs abnormal temperature rise confirmation signal;Execution module determines real abnormality, transient response or unstable state according to contact resistance change trend factor and load current fluctuation rate, and outputs early warning in stages.The present application realizes self-adaptive, multi-dimensional automatic inspection of terminal temperature rise, effectively distinguishes bad contact, overload and harmonic interference and other working conditions, reduces false alarm and missed alarm rate, and improves power distribution network operation safety and intelligent level.

Description

Technical Field

[0001] This invention belongs to the field of fault detection technology, and in particular relates to an automatic inspection system for the temperature rise of the wiring terminals of an electricity metering box. Background Technology

[0002] In existing solutions for detecting temperature rise at the terminals of power metering boxes, manual periodic inspections or point measurements using handheld infrared thermometers are typically relied upon. The entire process includes a series of steps such as on-site investigation, temperature data recording, threshold comparison, anomaly reporting, and subsequent handling. These steps together constitute a complete link for assessing the thermal condition of the terminals and preventing faults. The frequency and accuracy of this link directly affect the safe operation of the equipment and the reliability of the power system. Currently, such detection is mostly carried out manually or semi-automatically, and the temperature rise judgment logic is highly coupled with the specific metering box model, installation environment, load conditions, and the experience of maintenance personnel. Because metering boxes from different manufacturers, batches, and operating scenarios lack unified thermal characteristic standards in terms of terminal block structure, heat dissipation conditions, and rated current, their temperature rise patterns and alarm thresholds exhibit significant heterogeneity, ambiguity, and dynamic fluctuations. Systems often need to set separate detection rules and early warning strategies for each box type or scenario, resulting in fragmented diagnostic logic, redundant configuration, and difficulty in unified updates. Especially when distinguishing between subdivided states such as "localized overheating due to poor contact," "overall temperature rise caused by long-term overload," "false high temperatures due to ambient temperature influence," and "false alarms caused by instantaneous fluctuations," existing methods lack standardized temperature rise characteristic abstraction and configurable dynamic threshold management capabilities, making it impossible to flexibly adapt to new box types or adjust early warning rules. Furthermore, with the continuous increase in the number of connected metering boxes and the thermal characteristic drift caused by load changes and seasonal changes, temperature rise judgment criteria need frequent adjustments. Manual or hard-coded detection methods cannot adapt to these changes without modifying the detection strategy code and redeploying, resulting in high maintenance costs, delayed response, and a high risk of missed or false alarms due to improper adjustments. Meanwhile, due to the lack of a unified standardized temperature rise data sensing layer and a full-link inspection and tracking mechanism, when multiple metering boxes operate in parallel, the system struggles to monitor the real-time thermal evolution of each terminal, making fault location difficult. Furthermore, subsequent intelligent diagnostic models struggle to be trained and make decisions based on accurate temperature change trends. Therefore, how to achieve unified perception, dynamic semantic mapping, and configurable early warning management of temperature rise characteristics of heterogeneous metering box terminals through standardization, and how to build a reliable full-link automatic inspection and intelligent diagnostic mechanism based on this, has become a key issue in improving the operational safety of power metering boxes and ensuring the reliability of low-voltage distribution networks. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention proposes an automatic temperature rise inspection system for the terminals of an energy metering box. The system includes a data acquisition module to obtain real-time temperature, ambient temperature, and load current; a feature extraction module to generate a thermal state feature vector containing absolute temperature value, temperature rise rate, duration, and contact resistance change trend factor; an anomaly detection module to compare the feature vector with an adaptively updated dynamic threshold set and output an abnormal temperature rise confirmation signal; and an execution module to determine whether the anomaly is genuine, an instantaneous response, or an unstable state based on the contact resistance change trend factor and load current fluctuation rate, and to output tiered warnings. This invention achieves adaptive, multi-dimensional automatic inspection of terminal temperature rise, effectively distinguishing between poor contact, overload, and harmonic interference conditions, reducing false alarm and missed alarm rates, and improving the safety and intelligence level of power distribution network operation.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] An automatic temperature rise inspection system for the wiring terminals of an electricity metering box includes:

[0006] The data acquisition module is configured to acquire real-time temperature sequences, ambient temperature data, and load current data of the terminal blocks.

[0007] The feature extraction module is configured to: perform thermal state feature extraction based on the real-time temperature sequence, ambient temperature data, and load current data, and generate a thermal state feature vector. The thermal state feature vector includes absolute temperature value, temperature rise rate, duration, and contact resistance change trend factor. The temperature rise rate is calculated from the temperature difference and time interval between adjacent sampling times, and the contact resistance change trend factor is obtained by fitting the ratio of temperature change to load current change.

[0008] The anomaly detection module is configured to: compare the thermal state feature vector with a dynamic threshold set; and when it is detected that the absolute temperature value exceeds the first temperature threshold and the duration for which the temperature rise rate exceeds the second temperature rise rate threshold reaches the duration threshold, output an abnormal temperature rise confirmation signal; the dynamic threshold set includes a first temperature threshold, a second temperature rise rate threshold, and a duration threshold, and the dynamic threshold set is adaptively updated based on historical temperature statistics, ambient temperature compensation coefficient, and load current correction factor;

[0009] The execution module is configured to: execute an abnormal temperature rise confirmation signal.

[0010] Specifically, the abnormal temperature rise confirmation signal is executed, including:

[0011] When the abnormal temperature rise confirmation signal is received, the driver outputs the early warning information in a preset manner and obtains the online monitoring data under the abnormal state. The online monitoring data includes the real-time value of the contact resistance change trend factor and the fluctuation rate of the current load current.

[0012] If the contact resistance change trend factor is less than the third preset threshold and the fluctuation rate of the current load current is less than the fourth preset threshold, it is determined to be a real abnormal temperature rise. The synchronous control storage module records the current abnormal event and outputs a warning completion command.

[0013] Specifically, the execution of the abnormal temperature rise confirmation signal also includes:

[0014] If the contact resistance change trend factor is less than the third preset threshold, but the fluctuation rate of the current load current is greater than or equal to the fourth preset threshold, it is determined to be an instantaneous temperature response caused by a sudden load change. The abnormal event is not recorded for the time being. The current temperature sequence is maintained and the abnormal judgment module is re-executed after a delay of one sampling period.

[0015] If the contact resistance change trend factor is greater than or equal to the third preset threshold, the retry counter is started. When the retry count is less than the fifth preset threshold, the current temperature sequence is cleared and the thermal state feature extraction step is returned to be executed. At the same time, the retry count is incremented by one. When the retry count reaches the fifth preset threshold, it is determined to be a continuous unstable contact state. A warning completion command is forcibly output and the retry counter is reset.

[0016] Specifically, the feature extraction module includes:

[0017] The first preprocessing unit is configured to perform adaptive filtering on the real-time temperature sequence of the terminals to obtain a smooth temperature sequence; wherein the adaptive filtering adopts a series structure of sliding window mid-range filtering and first-order hysteresis filtering; the window length L of the sliding window mid-range filtering is based on the load current fluctuation rate σ. I The adjustments are as follows:

[0018] σ I It equals the ratio of the standard deviation to the mean of the load current sequence within the current sliding window; when σ I Greater than the preset fluctuation threshold δ I When σ is reached, L is set as the first length L1; when σ is reached... I Less than or equal to δ I When L is set to the second length L2, and L1 > L2;

[0019] The filter coefficients of the first-order hysteresis filter It is calculated adaptively based on the local variance of the temperature series, specifically as follows: ,in For the estimated noise variance, This represents the variance of the local signal change over k temperature values ​​before and after the current sampling point.

[0020] Specifically, the feature extraction module further includes:

[0021] The second phase compensation unit is configured to: receive the smoothed temperature sequence and load current data, perform phase compensation on the load current data, and obtain an equivalent excitation current sequence; the equivalent excitation current sequence and the smoothed temperature sequence are aligned in phase, such that the absolute value of the phase difference between the theoretical response temperature after the equivalent excitation current is subjected to the terminal thermal inertia transfer function and the smoothed temperature sequence is less than a preset phase difference threshold.

[0022] The third fundamental frequency separation unit is configured to: receive the equivalent excitation current sequence, perform fundamental frequency separation processing on the equivalent excitation current sequence to obtain the effective value sequence of the fundamental frequency current, and filter out the harmonic components in the equivalent excitation current sequence.

[0023] Specifically, the anomaly detection module includes: a first threshold adaptive unit, configured to generate a dynamic threshold set for the current moment based on historical temperature statistics, ambient temperature compensation coefficient, and load current correction factor; the dynamic threshold set includes a first temperature threshold, a second temperature rise rate threshold, and a duration threshold.

[0024] Specifically, the anomaly detection module further includes: a second condition determination unit, configured to: perform multi-condition joint determination based on the final thermal state feature vector and the dynamic threshold set; when the determination result meets the abnormal temperature rise condition, output the monitoring point identifier that meets the condition and the corresponding thermal state feature vector to the third confirmation signal generation unit, specifically:

[0025] E1. Extract the absolute temperature value, temperature rise rate, and duration of each monitoring point from the final thermal state feature vector; wherein, the duration refers to the cumulative length of time during which the absolute temperature value of the monitoring point continuously exceeds the first temperature threshold, and the temperature rise rate simultaneously exceeds the second temperature rise rate threshold during the cumulative time.

[0026] E2. For each monitoring point, determine whether the following three conditions are met simultaneously:

[0027] Condition 1: The absolute value of the current temperature is greater than the first temperature threshold.

[0028] Condition 2: The current temperature rise rate is greater than the second temperature rise rate threshold;

[0029] Condition 3: The current duration is greater than or equal to the duration threshold.

[0030] E3. If the current monitoring point meets all three judgment conditions at the same time, it is determined that the monitoring point meets the abnormal temperature rise condition, and the identifier of the monitoring point and the thermal state feature vector corresponding to the current time are output to the third confirmation signal generation unit; if it does not meet the condition, the monitoring data of the next sampling time are judged.

[0031] Specifically, the execution of the abnormal temperature rise confirmation signal also includes an abnormality level classification step, which is as follows:

[0032] When a true abnormal temperature rise is determined, the difference between the absolute value of the current temperature and the first temperature threshold is obtained, and the temperature exceedance ratio is calculated. The temperature exceedance ratio is equal to the difference divided by the first temperature threshold.

[0033] When the temperature exceedance ratio is less than the first-level threshold, the abnormality level is marked as a first-level abnormality, and the warning output module is controlled to output a yellow warning message.

[0034] When the temperature exceedance ratio is between the first-level threshold and the second-level threshold, the abnormality level is marked as a level two abnormality, an orange warning message is output, and on-site inspection is recommended.

[0035] When the temperature exceedance ratio is greater than or equal to the second-level threshold, the abnormality level is marked as a third-level abnormality, a red warning message is output, and a remote power-off protection command is triggered.

[0036] Specifically, the execution of the abnormal temperature rise confirmation signal also includes a step of retrying the historical cumulative record, as follows:

[0037] An independent cumulative instability counter is maintained for each monitoring point. This counter is used to record the number of times the monitoring point reaches the fifth preset threshold and triggers a mandatory warning within a preset time window.

[0038] When the retry count reaches the fifth preset threshold and a warning completion command is forcibly output, the accumulated instability counter is incremented by one, and the current timestamp is recorded.

[0039] When the accumulated instability counter exceeds the sixth preset threshold, it is determined that the monitoring point has a chronic contact degradation trend, and a chronic degradation early warning message is generated and pushed to the operation and maintenance platform.

[0040] Specifically, when the anomaly detection module is re-executed after a one-sampling-period delay, a conflict handling process during the delay period is also included, specifically:

[0041] If an abnormal temperature rise confirmation signal is received again from the same monitoring point during the delay period, the signal is ignored and the judgment is not repeated.

[0042] After the delay ends, when re-entering the anomaly detection module, the temporary status flags accumulated during the delay period are cleared, and the latest temperature sequence and load current data are used as input.

[0043] If the number of consecutive ignored abnormal temperature rise confirmation signals exceeds the seventh preset threshold during the delay period, it will be directly determined as a real abnormal temperature rise, and the remaining discrimination will be skipped to directly output an early warning.

[0044] Compared with the prior art, the beneficial effects of the present invention are:

[0045] This invention acquires real-time temperature, ambient temperature, and load current data of the terminal blocks through a data acquisition module. A feature extraction module generates a thermal state feature vector containing absolute temperature value, temperature rise rate, duration, and contact resistance change trend factors, achieving a comprehensive and refined characterization of the terminal block's thermal state and overcoming the limitations of traditional single-threshold detection. The anomaly detection module compares data against a dynamically updated set of thresholds based on historical data, ambient temperature, and load current, dynamically adapting to different operating conditions and environmental changes, avoiding missed or false alarms caused by fixed thresholds. The execution module further performs secondary discrimination on the abnormal temperature rise confirmation signal based on the contact resistance change trend factor and load current fluctuation rate, distinguishing between genuine abnormal temperature rises, instantaneous responses caused by load mutations, and continuously unstable contact states. A retry counter mechanism is introduced to effectively eliminate false alarms caused by instantaneous disturbances and electromagnetic interference, significantly reducing the false alarm rate. Simultaneously, multi-level early warning and forced output logic ensure reliable reporting of genuine faults. In summary, this invention achieves automatic, intelligent, and highly reliable inspection of terminal block temperature rises, improving the operational safety of power metering boxes and the intelligent maintenance level of the distribution network. Attached Figure Description

[0046] Figure 1 This is a module unit diagram of an automatic inspection system for temperature rise of wiring terminals in an energy metering box according to an embodiment of the present invention;

[0047] Figure 2 This is a flowchart of the execution module of the present invention. Detailed Implementation

[0048] Example 1

[0049] Please see Figure 1 One embodiment of the present invention provides an automatic temperature rise inspection system for the wiring terminals of an electricity metering box, comprising:

[0050] The data acquisition module is configured to acquire real-time temperature sequences, ambient temperature data, and load current data of the terminal blocks.

[0051] The feature extraction module is configured to: perform thermal state feature extraction based on the real-time temperature sequence, ambient temperature data, and load current data, and generate a thermal state feature vector. The thermal state feature vector includes absolute temperature value, temperature rise rate, duration, and contact resistance change trend factor. The temperature rise rate is calculated from the temperature difference and time interval between adjacent sampling times, and the contact resistance change trend factor is obtained by fitting the ratio of temperature change to load current change.

[0052] It should be further explained that the feature extraction module of this embodiment includes: a first preprocessing unit, a second phase compensation unit, a third fundamental wave separation unit, a fourth spatial feature extraction unit, and a fifth trend factor fitting unit;

[0053] It should be further explained that the first preprocessing unit in this embodiment is configured to: perform adaptive filtering on the real-time temperature sequence of the terminals to obtain a smooth temperature sequence; wherein, the adaptive filtering adopts a series structure of sliding window mid-range filtering and first-order hysteresis filtering; the window length L of the sliding window mid-range filtering is based on the load current fluctuation rate σ I The adjustments are as follows:

[0054] σ I It equals the ratio of the standard deviation to the mean of the load current sequence within the current sliding window; when σ I Greater than the preset fluctuation threshold δ I When σ is reached, L is set as the first length L1; when σ is reached... I Less than or equal to δ I When L is set to the second length L2, and L1 > L2; in this embodiment, the preset fluctuation threshold δ is used. I The specific values ​​of the first length L1 and the second length L2 can be determined through offline testing based on the rated current range of the terminal block, historical load fluctuation statistics, and temperature acquisition noise characteristics; typically, δ I It can be set between 0.15 and 0.25 to distinguish between stable loads and impulsive loads, when the current fluctuation rate σ I Greater than δ I To effectively suppress the instantaneous temperature spikes introduced by the impact current, the window length L is set to a relatively large L1 (e.g., 7 or 9). When σ I Less than or equal to δ I In order to ensure the real-time performance of temperature tracking, L is set to a smaller value, L2 (such as 3 or 5). The values ​​of the parameters in this embodiment can be routinely adjusted by those skilled in the art based on actual operating conditions, dynamic characteristics of the metering box, and inspection sensitivity requirements, without departing from the protection scope of this invention.

[0055] The filter coefficients of the first-order hysteresis filter It is calculated adaptively based on the local variance of the temperature series, specifically as follows: ,in For the estimated noise variance, This represents the variance of the local signal change over k temperature values ​​before and after the current sampling point.

[0056] It should be further noted that the second phase compensation unit in this embodiment is configured to: receive the smoothed temperature sequence T smooth (t) and load current data, phase compensation is performed on the load current data to obtain an equivalent excitation current sequence; the equivalent excitation current sequence and the smoothed temperature sequence are phase-aligned, such that the absolute value of the phase difference between the theoretical response temperature after the equivalent excitation current is applied by the terminal thermal inertia transfer function and the smoothed temperature sequence is less than a preset phase difference threshold; in this embodiment, the setting of the preset phase difference threshold needs to be comprehensively determined in combination with the calibration accuracy of the terminal thermal inertia time constant, the temperature acquisition noise level, and the system's real-time requirements for abnormal temperature rise detection; typically, it can be based on the phase difference between the theoretical response temperature and the smoothed temperature sequence obtained from multiple repeated measurements in historical step response tests. The maximum fluctuation range is taken as the 97% confidence upper limit as the threshold benchmark, and the additional time delay introduced by the field load fluctuation is considered. On this basis, a margin of 10% to 20% is reserved. The typical preset phase difference threshold can be set between 5° and 10° (corresponding to a delay of 0.28ms to 0.56ms at a power frequency of 50Hz). This value can effectively tolerate the small phase mismatch caused by the simplification of the thermal inertia model and the time-varying parameters, and can also avoid the distortion of the contact resistance change trend factor calculation due to phase alignment failure. The preset phase difference threshold can be routinely adjusted by those skilled in the art according to the actual thermal parameter dispersion of the equipment and the inspection sensitivity requirements, without departing from the protection scope of this invention.

[0057] The terminal thermal inertia transfer function is a first-order inertial element: G(s) = 1 / (τ·s+1). The time constant τ is pre-calibrated by historical step response experiments. The calibration method is as follows: apply a step load current to the terminal, record the temperature response curve, and determine the initial value of τ as the time required for the temperature response to reach 63.2% of the steady-state value. In automatic control theory and transfer function expressions, s is a complex variable (i.e., the Laplace operator), which is located in the complex frequency domain and essentially represents differential operations. In the terminal thermal inertia model G(s) = 1 / (τ·s+1) of this application, the differential equation in the time domain is transformed into an algebraic expression in the complex frequency domain, simplifying the originally complex dynamic response calculation into an intuitive first-order inertial element. The product of s and the time constant τ reflects the hysteresis effect of the system on input changes (such as sudden changes in load current). Physically, it is used to characterize the transient temperature rise response characteristics of the terminal from current change to temperature stability, which is determined by thermal capacity and thermal resistance. It is the mathematical basis for realizing phase compensation and current-temperature timing alignment.

[0058] It should be further explained that, in this embodiment, phase compensation specifically adopts a combination of inverse filtering and error iterative correction, as follows:

[0059] A1, the original load current data is processed through the inverse transfer function G -1 (s) Perform inverse filtering to obtain the initial equivalent excitation current sequence I. eq0 (t);

[0060] A2, I eq0 Substituting (t) into the forward transfer function G(s), the theoretical temperature sequence T is calculated. theory0 (t);

[0061] A3, Calculate T theory0 (t) and T smooth The root mean square error RMSE0 between (t) is less than the preset error threshold ε. RMSE Then directly output I. eq0 (t) serves as the equivalent excitation current sequence;

[0062] A4, if RMSE0 is greater than or equal to ε RMSE Then, the adaptive correction of the time constant τ is initiated, specifically as follows: taking the current τ as the center, traverse the interval [0.5τ, 2τ] with a step size Δτ, calculate the inverse filtering result and theoretical temperature sequence corresponding to each τ value, and select the τ that minimizes RMSE0. opt As the corrected time constant; in this embodiment, a preset error threshold ε is used. RMSE The settings need to comprehensively consider the steady-state noise level of the smoothed temperature series, the upper limit of the calibration error of the time constant τ of the terminal thermal inertia, and the allowable deviation of the phase compensation in calculating the contact resistance variation trend factor. Typically, the standard deviation σ of the smoothed temperature series fluctuation can be extracted based on long-term steady-state operating data of the terminals under rated current. 稳态 Then, combining the theoretical response residual caused by the simplification of the transfer function model (first-order inertial element), ε RMSE Initialized to max(3σ) 稳态 Rated temperature difference ΔT 额定 2%), where ΔT 额定 Take the steady-state temperature rise value of the terminal when it carries the rated current; if the thermal parameters drift due to terminal aging or drastic environmental changes, the operation and maintenance platform can automatically verify and update ε through batch processing of historical data. RMSE .

[0063] A5, using τ opt Re-execute the inverse filter to obtain the final equivalent excitation current sequence I. eq(t); or a state observer approach can be used, specifically: constructing a state-space model with the state equation dI. eq (t) / dt=0, meaning it is assumed that the equivalent excitation current is constant within the preset sampling interval, and the observation equation is: Where t represents the current time, This represents the integral variable, i.e., a specific historical moment in the past. This is the smoothed temperature value at the current time t, i.e., the theoretical temperature output after thermal inertia filtering. For a historic moment The equivalent excitation current value (i.e., the load current after phase compensation) is given by τ, which is the time constant of the terminal thermal inertia, reflecting the speed of temperature response to current changes. Numerically, it is equal to the time required for the temperature response to rise from the initial value to 63.2% of the steady-state value. The weighting function is an exponentially decaying function, representing historical time points. The effect of current on current temperature with time interval The sampling interval decreases exponentially as the temperature rises. In this embodiment, the "preset sampling interval" refers to the system's acquisition period for temperature sequence and load current data. Its setting needs to comprehensively consider the thermal inertia time constant τ of the terminal, the typical spectrum of load current fluctuations, and the resolution requirements for abnormal temperature rise rate detection. Generally, the sampling interval should meet the conditions for accurate calculation of the temperature rise rate, ensuring that the temperature change within two adjacent sampling intervals is significantly greater than the temperature measurement noise. At the same time, in order to capture the instantaneous thermal response caused by load mutations, the sampling interval should preferably be selected as 1 / 5 to 1 / 10 of the terminal thermal inertia time constant (if τ is on the order of minutes, the sampling interval can be set to several seconds to tens of seconds), and should be much lower than the baseline value of the duration threshold to ensure the precision of the duration statistics. In practical engineering applications, this sampling interval can be routinely adjusted by those skilled in the art according to the thermal capacity characteristics of the metering box, the severity of on-site load fluctuations, and the requirements for inspection sensitivity, all without departing from the protection scope of this invention.

[0064] Using the smoothed temperature sequence as the observed value and the equivalent excitation current as the state variable to be estimated, the estimated value of the equivalent excitation current at the current moment is calculated recursively through a Kalman filter. The process noise covariance Q and the observation noise covariance R of the Kalman filter are pre-calibrated according to the fluctuation variance of the historical temperature sequence.

[0065] After compensation is completed, verify the final equivalent excitation current sequence I. eq (t) The theoretical temperature sequence calculated using the transfer function G(s) and T smooth Is the root mean square error between (t) less than ε? RMSEThe system checks whether the maximum single-point error is less than a preset single-point error threshold. If both conditions are met, the phase compensation is confirmed to be effective. If any condition is not met, a phase compensation failure flag is output, and the uncompensated original load current data is used as a substitute for the equivalent excitation current sequence. In this embodiment, during the phase compensation verification stage, the final equivalent excitation current sequence I is used. eq (t) Input terminal thermal inertia transfer function G(s)=1 / (τ·s+1), calculate the theoretical temperature value at each sampling time, and compare it with the smoothed temperature sequence T at the corresponding time. smooth (t) Take the absolute value of the difference, iterate through all sampling points in the entire verification time period, and take the maximum value of the absolute value of the above difference as the maximum single-point error.

[0066] In this embodiment, the preset single-point error threshold is used to constrain the maximum permissible deviation between the theoretical response temperature and the measured smooth temperature of a single sampling point. Its value needs to comprehensively consider the measurement accuracy level of the temperature sensor, the upper limit of the simplified residual of the first-order inertial model for the actual heat conduction process, and the amplitude of instantaneous temperature spikes that may be introduced by local electromagnetic interference. Typically, the standard deviation σ of the temperature measurement noise during steady-state operation is used. 噪声 Based on this, the preset single-point error threshold can be set to 5σ. 噪声 Up to 8σ 噪声 Alternatively, take the rated temperature rise ΔT of the wiring terminal. 额定 3% to 5% and the above multiple σ 噪声 The larger of the two values ​​is chosen so that it can tolerate the normal fitting error caused by the linearization approximation of the model, and can effectively detect local calculation divergence caused by inaccurate τ calibration or failure of phase compensation. The specific value of the above threshold can be routinely adjusted by those skilled in the art based on the actual temperature acquisition accuracy of the metering box, the dispersion of thermal inertia parameters and the tolerance of false alarms during inspection, without departing from the protection scope of this invention.

[0067] The confirmed effective equivalent excitation current sequence and the smoothed temperature sequence are synchronously output to the third fundamental wave separation unit.

[0068] It should be further explained that the third fundamental frequency separation unit in this embodiment is configured to: receive the equivalent excitation current sequence, perform fundamental frequency separation processing on the equivalent excitation current sequence to obtain the fundamental frequency current effective value sequence, and filter out the harmonic components in the equivalent excitation current sequence;

[0069] It should be further noted that, in this embodiment, one implementation of the fundamental wave separation process includes:

[0070] B1. The three-phase equivalent excitation current is converted into α and β components in a two-phase stationary coordinate system through Clark transformation. The Clark transformation adopts an equal amplitude transformation method. The transformed α component is equal to the A phase current minus 1 / 2 of the sum of the B phase current and the C phase current, and then multiplied by 2 / 3. The β component is equal to the square root of 2 / 3 multiplied by the difference between the B phase current and the C phase current, and then multiplied by 2 / 3.

[0071] B2. The α and β components are input into a digital phase-locked loop (PLL), which is composed of a phase detector, a loop filter, and a voltage-controlled oscillator (VCO) connected in series. The phase detector receives the α and β components and calculates the current phase error. Specifically, it calculates the ratio of the β component to the α component, takes the arctangent of this ratio to obtain the phase angle of the current voltage vector, and then subtracts the estimated phase output by the PLL. The difference is the phase error. The loop filter uses a proportional-integral (PI) controller. It multiplies the current phase error by a proportional coefficient to obtain a proportional term, integrates the phase error, and multiplies it by an integral coefficient to obtain an integral term. The proportional term and the integral term are added to obtain the frequency correction. The VCO adds the frequency correction based on the estimated frequency from the previous moment and then integrates it to obtain the estimated phase at the current moment.

[0072] When the absolute value of the phase error is less than the preset phase-locked loop (PLL) threshold for N consecutive sampling periods, the PLL is determined to be locked, and the fundamental frequency and synchronization phase of the PLL are output. If the locking condition is not met after a preset timeout period from the start of the PLL, a fundamental separation failure flag is output, and subsequent fundamental separation steps are abandoned, with the effective value of the total current used as the alternative output. In this embodiment, the preset PLL threshold is used to determine whether the PLL has reliably locked the fundamental frequency and phase, and its value must take into account the grid frequency stability and current harmonic distortion. The degree of variation and the bandwidth of the phase-locked loop (PLL) loop filter; typically, the peak value of the steady-state phase error when the PLL is stably tracking under rated operating conditions is used as a benchmark, and 1.5 to 2 times this benchmark value is taken as the preset phase-locked threshold (if the power grid frequency fluctuation is less than ±0.1Hz and the total harmonic distortion rate is less than 5%, the typical phase error can be controlled within 2° to 4°, then the preset phase-locked threshold can be set to 5° to 8°). When the absolute value of the phase error is lower than this threshold for N consecutive sampling periods, it is considered that the PLL has left the transient acquisition process and entered the high-precision synchronization state.

[0073] In this embodiment, the preset timeout is used to prevent the phase-locked loop (PLL) from waiting indefinitely to lock due to severe grid distortion, signal interruption, or hardware malfunction. Its value should be greater than the maximum time required for the PLL to complete capture under adverse but normal operating conditions. Typically, it can be calibrated based on the maximum phase-locked recovery time under field startup or load switching impact. The time from PLL startup to the first drop of phase error to within the preset phase-locked threshold under various operating conditions is measured and recorded. The 98th percentile of this set of time data is taken and an additional safety margin of 1 to 2 seconds is added as the preset timeout (typically, it can be set between 2 and 5 seconds).

[0074] In this embodiment, the values ​​of the preset phase-locked threshold and the preset timeout can be routinely adjusted by those skilled in the art based on the actual power grid quality, current sampling accuracy, and the inspection system's tolerance for real-time performance, without departing from the protection scope of this invention.

[0075] B3. After the phase-locked loop (PLL) is locked, the α and β components are transformed to a rotating coordinate system using the Parker transformation to obtain the direct-axis component and the quadrature-axis component. The Parker transformation uses the synchronization phase output by the PLL as the rotation angle. Specifically, the direct-axis component is equal to the α component multiplied by the cosine of the synchronization phase plus the β component multiplied by the sine of the synchronization phase; the quadrature-axis component is equal to the negative α component multiplied by the sine of the synchronization phase plus the β component multiplied by the cosine of the synchronization phase.

[0076] B4 performs low-pass filtering on the direct-axis and quadrature-axis components respectively to remove the AC components and obtain the direct-axis DC components and quadrature-axis DC components. The low-pass filter adopts a second-order Butterworth type, and the cutoff frequency is set to half of the fundamental frequency. The output value after filtering is the DC component of each component.

[0077] B5. Calculate the effective value of the fundamental current based on the direct-axis DC component and the quadrature-axis DC component. Specifically, calculate the sum of the squares of the direct-axis DC component and the quadrature-axis DC component, and then take the square root of the sum. The resulting value is the effective value of the fundamental current at the current sampling time.

[0078] B6. Calculate the change in the fundamental current effective value, which is the fundamental current effective value at the current sampling time minus the fundamental current effective value at the previous sampling time; at the same time, calculate the change in the total current effective value, which is the total current effective value at the current sampling time minus the total current effective value at the previous sampling time; the total current effective value is obtained by dividing the sum of the squares of the three-phase equivalent excitation currents by 3 and then taking the square root.

[0079] The change in the fundamental current effective value is compared with the change in the total current effective value. Specifically, the absolute value of the difference between the change in the fundamental current effective value and the change in the total current effective value is calculated first, and then divided by the sum of the absolute value of the change in the total current effective value and a preset small positive number to obtain the ratio. When the ratio is greater than the preset harmonic interference threshold, it is determined that there is significant harmonic distortion at the current sampling point, and a harmonic interference flag is generated with a true value; otherwise, the flag is false. In this embodiment, the preset small positive number is used to prevent overflow of the division operation when the change in the total current effective value is zero. Its value should be much smaller than the minimum resolvable change in the total current effective value within a single sampling interval when the terminal is operating normally. Typically, it can be set as 1 / 10 to 1 / 5 of the minimum resolution value of the current transformer or sampling circuit as the preset small positive number. A typical value can be set to 0.01A to 0.05A, which can avoid the calculation abnormality of the denominator being zero, and will not affect the sensitivity of the ratio to small current fluctuations.

[0080] In this embodiment, a preset harmonic interference threshold is used to distinguish the difference in current change caused by normal load fluctuations and significant harmonic distortion. Its value is determined based on the following: when there are no significant harmonics in the power grid, the change in the sampling interval of the fundamental current RMS value should be close to the change in the total current RMS value, with the difference mainly contributed by measurement noise and minor interharmonics; when significant harmonic distortion exists, the proportion of harmonic components in the total current increases, causing the change in the total current RMS value to deviate significantly from the change in the fundamental current RMS value. Therefore, historical operating data of the terminals under known harmonic interference (or total harmonic distortion rate less than 1%) can be collected, the statistical distribution of the above ratios at each sampling point can be calculated, and the 95th percentile or three times the upper limit of the standard deviation can be taken as the preset harmonic interference threshold. A typical value can be set between 0.15 and 0.30.

[0081] In this embodiment, the specific values ​​of the preset small positive number and the preset harmonic interference threshold can be routinely adjusted by those skilled in the art based on the actual current sampling accuracy, the harmonic background level of the distribution network where the terminal is located, and the sensitivity requirements of the inspection system for harmonic distortion detection, all without departing from the protection scope of this invention.

[0082] B7 replaces the change in the effective value of the fundamental current with the change in the effective value of the total current, and outputs it along with the harmonic interference flag to the fourth spatial feature extraction unit.

[0083] It should be further explained that the third fundamental wave separation unit in this embodiment is further configured to perform the following technical processes:

[0084] Firstly, in the phase-locked loop (PLL) locking state determination stage, the preferred value range for the number of consecutive sampling periods N is 5 to 10. The value of N is used to determine the stable locking state of the PLL. The PLL is determined to be locked only when the absolute value of the phase error is less than the preset PLL threshold for N consecutive sampling periods, so as to avoid false locking caused by instantaneous phase error reaching the standard. At the same time, an adaptive adjustment logic for the value of N is configured. When the amplitude of the power grid frequency fluctuation is detected to exceed the preset frequency fluctuation threshold, the value of N is adaptively increased to improve the anti-interference and reliability of the locking determination. In this embodiment, the preferred range for the number of continuous sampling periods N is set to 5 to 10. The rationale for this preference is as follows: Under non-ideal operating conditions where the power grid frequency fluctuates slightly, current harmonic distortion and measurement noise coexist, the instantaneous phase error value of the phase-locked loop may accidentally fall within the preset phase-locked threshold due to random disturbances. If the lock is determined based on a single or a few consecutive fulfillments of the condition, false locks are very likely to occur, leading to a shift in the subsequent fundamental phase separation reference. When N is too small (e.g., 1 to 3 periods), the ability to resist random disturbances is insufficient. When N is too large (e.g., more than 20 periods), although the confidence of the lock determination can be enhanced, it will significantly prolong the total time from start-up to lock confirmation of the phase-locked loop, reducing the system's real-time tracking performance for rapid load changes. Selecting 5 to 10 sampling periods as the continuous determination window can effectively filter out false locks caused by instantaneous disturbances in more than 95% of normal situations, while controlling the lock confirmation delay to within a few times the sampling period, thus balancing the reliability and speed of phase-locked determination. The specific value of this range can be routinely fine-tuned by those skilled in the art based on the actual power grid frequency stability, sampling frequency, and system real-time requirements, without departing from the protection scope of this invention. In this embodiment, the setting of the preset frequency fluctuation threshold needs to be determined based on the rated frequency of the distribution network where the terminal is located and its allowable steady-state frequency deviation range during normal operation. Its value should be able to effectively distinguish between the inherent small frequency drift of the power grid and the significant frequency disturbances caused by drastic load switching, short-circuit faults, or sudden changes in power generation. Typically, based on national or industry standards for power grid frequency quality (e.g., the normal frequency fluctuation limit for a 50Hz system is specified as ±0.2Hz), the frequency data of the access point during historical normal operation periods can be continuously monitored, the standard deviation σf of the frequency sequence can be calculated, and the larger of 3σf and the standard limit can be taken as the preset frequency fluctuation threshold. A typical value can be set between 0.2Hz and 0.5Hz. When the instantaneous amplitude of the power grid frequency fluctuation exceeds the threshold, it is determined that there is an abnormal fluctuation in the current frequency environment. In this case, the number of continuous sampling periods N of the phase-locked loop needs to be increased accordingly to enhance the anti-interference capability of the locking determination. Conversely, when the frequency fluctuation amplitude is less than or equal to the threshold, the normal value of N is maintained. In this embodiment, the frequency fluctuation threshold is routinely adjusted by those skilled in the art based on the actual power grid frequency stability level and the trade-off requirements of the inspection system for phase-locking speed and locking reliability, all of which do not depart from the protection scope of this invention.

[0085] Secondly, the full-link linkage application logic for configuring harmonic interference markers is as follows: the harmonic interference markers output by the third fundamental wave separation unit are synchronously transmitted to the fourth spatial feature extraction unit, the anomaly discrimination module, and the execution module; the fourth spatial feature extraction unit is configured to, when receiving a harmonic interference marker with a true value, reduce the calculation weight of the temperature data corresponding to the current sampling point by 50% during the calculation of the global maximum value, global average value, and global variance, to suppress the interference of temperature calculation deviation caused by harmonic distortion on spatial distribution feature extraction; the anomaly discrimination module is configured to, when receiving a harmonic interference marker with a true value, temporarily increase the second temperature rise rate threshold at the current moment by 20%, to avoid misjudgment of temperature rise rate caused by distortion of the effective current value due to harmonics; the execution module is configured to, when the duration of the detected harmonic interference marker being true exceeds a preset duration threshold, use harmonic distortion as an auxiliary anomaly cause and output it synchronously to the operation and maintenance platform along with the anomaly temperature rise confirmation signal. In this embodiment, the setting rules for the calculation weights are further clarified as follows: Under normal operating conditions where no harmonic interference markers are detected, the weights of the temperature data at each monitoring point participating in the calculation of the global maximum value, global average value, and global variance are all defaulted to 1.0 (i.e., equal weights are used in the statistics); when the third fundamental wave separation unit outputs a harmonic interference marker with a true marker value, the fourth spatial feature extraction unit reduces the calculation weight of the temperature data corresponding to the current sampling point by 50% from the default value of 1.0 to 0.5, so as to reduce the statistical contribution of the temperature data at that moment. The basis for reducing the weight by 50% is that harmonic distortion causes the effective value of the total current to be distorted, and the resulting temperature measurement deviation is not completely invalid. Compared with directly eliminating the sampling point (resetting the weight to zero), the half-weighting can suppress the pollution of spatial distribution feature extraction by harmonic interference to a certain extent, and can also retain some of the real thermal state information still contained in the sampling point. Through actual measurement and comparison, within the typical harmonic total distortion rate (5%~15%) of the terminal block, a 50% weight attenuation can effectively reduce the calculation deviation of the spatial correlation coefficient, while avoiding the omission of global extreme values ​​due to excessive suppression. This ratio can be routinely adjusted by those skilled in the art according to the actual harmonic level of the distribution network and the inspection sensitivity requirements, without departing from the protection scope of this invention.

[0086] Third, a hard synchronization execution mechanism for Clark transformation, Park transformation, and current sampling is configured. Specifically, the triggering time of Clark transformation, digital phase-locked loop operation, and Park transformation is strictly bound to the completion time of three-phase current sampling. After each sampling cycle of three-phase equivalent excitation current sampling is completed, Clark transformation, digital phase-locked loop phase tracking operation, and Park transformation are immediately triggered to ensure that the direct-axis component and quadrature-axis component of the transformation output maintain a one-to-one time correspondence with the current sampling data of the current sampling cycle, eliminating the calculation error of the fundamental current effective value caused by the lag of the transformation operation.

[0087] It should be further explained that the fourth spatial feature extraction unit in this embodiment is configured as follows:

[0088] C1, the smoothed temperature values ​​of each monitoring point are arranged according to a preset monitoring point order to form a temperature absolute value vector. Each component of this vector corresponds to the temperature value of a monitoring point. In this embodiment, the preset monitoring point order refers to the spatial position numbering arrangement rule of each terminal temperature sensor in the power metering box. The setting is based on the actual physical layout of the terminals inside the metering box, starting from the incoming side and ending at the outgoing side, or arranged in the phase sequence of A phase, B phase, C phase, N phase. Each monitoring point is uniquely identified and numbered, and the numbering order is fixed. Once this order is set, it remains unchanged throughout the entire inspection cycle, so that the position of each component in the temperature absolute value vector corresponds one-to-one with the physical terminals. This ensures that when the fourth spatial feature extraction unit calculates the spatial correlation coefficient between the temperature absolute value vector and the standard load distribution vector, the correspondence of the vector elements is accurate and can truly reflect the matching degree of the temperature distribution of each monitoring point with the load distribution ratio, effectively distinguishing the spatial thermal characteristic differences between local single-phase contact failure and three-phase overall overload. The numbering rules for the above-mentioned monitoring points can be routinely adjusted by those skilled in the art based on the specific metering box structure, phase sequence definition, and operation and maintenance management habits, without departing from the protection scope of this invention.

[0089] C2, For each monitoring point, calculate the difference between the smoothed temperature value at the current sampling time and the smoothed temperature value at the previous sampling time, divide it by the sampling time interval, and obtain the temperature rise rate of the monitoring point; arrange the temperature rise rates of all monitoring points in the same order as the absolute temperature value vector to form a temperature rise rate vector;

[0090] C3, a standard load distribution vector is pre-set according to the rated current distribution ratio of each monitoring point. Each component of the vector is equal to the rated current of the corresponding monitoring point divided by the sum of the rated currents of all monitoring points, and the sum of all components is 1.

[0091] C4. Calculate the Pearson correlation coefficient between the absolute temperature vector and the standard load distribution vector as the first spatial correlation coefficient; calculate the Pearson correlation coefficient between the temperature rise rate vector and the standard load distribution vector as the second spatial correlation coefficient; the Pearson correlation coefficient reflects the degree of linear correlation between the two vectors and is equal to the product of the covariance of the two vectors and their respective standard deviations.

[0092] C5, determine the maximum value in the absolute temperature vector as the global maximum value; calculate the arithmetic mean of all components in the absolute temperature vector as the global average value; calculate the average of the sum of squares of the differences between each component in the absolute temperature vector and the global average value as the global variance.

[0093] C6 combines the absolute temperature value of each monitoring point, the temperature rise rate of each monitoring point, the first spatial correlation coefficient, the second spatial correlation coefficient, the global maximum value, the global average value, and the global variance into a thermal state feature vector, and outputs it to the fifth trend factor fitting unit.

[0094] It should be further explained that the fifth trend factor fitting unit in this embodiment is configured as follows:

[0095] The system receives the thermal state feature vector output by the fourth spatial feature extraction unit and the change in the effective value of the fundamental current output by the third fundamental current separation unit; wherein, the thermal state feature vector contains the absolute temperature value of each monitoring point at each sampling time;

[0096] For each monitoring point, the absolute temperature value of the monitoring point at multiple consecutive sampling times is extracted from the thermal state feature vector, and the temperature change at adjacent sampling times is calculated to obtain the temperature change sequence of the monitoring point.

[0097] Extract the sequence of changes in the fundamental current effective value of the corresponding monitoring point within the same time period from the changes in the fundamental current effective value;

[0098] The temperature change sequence and the fundamental current effective value change sequence are subjected to least squares linear fitting, and the linear regression coefficient obtained by fitting is used as the contact resistance change trend factor of the monitoring point.

[0099] The contact resistance change trend factor of each monitoring point is incorporated into the thermal state feature vector, and together with the original absolute temperature value of each monitoring point, temperature rise rate of each monitoring point, first spatial correlation coefficient, second spatial correlation coefficient, global maximum value, global average value, and global variance, they constitute the final thermal state feature vector.

[0100] The final thermal state feature vector is output to the anomaly detection module.

[0101] In this embodiment, the first preprocessing unit effectively suppresses the interference of high-frequency random noise and instantaneous spike pulses on the temperature rise rate calculation under strong electromagnetic environment by cascading adaptive sliding window mid-value filtering and first-order hysteresis filtering, so that the smooth temperature sequence truly reflects the heat accumulation process of the terminal; the second phase compensation unit, based on the transfer function model of the terminal thermal inertia, uses inverse filtering and error iterative correction or Kalman filtering to perform phase lead compensation on the load current, eliminating the phase mismatch caused by the temperature response lags behind the current change, and ensuring the timing alignment of the current and temperature changes in the contact resistance change trend factor; the third fundamental wave separation unit uses a digital phase-locked loop and Parker transform to extract the effective value of the fundamental current and filter out harmonic components, avoiding the distortion of the total effective value of the contact resistance caused by harmonic distortion. The fourth spatial feature extraction unit constructs a temperature absolute value vector and a temperature rise rate vector, calculates their spatial correlation coefficient with the standard load distribution vector and global statistics, effectively distinguishing between local single-phase contact failures and three-phase overall overload, preventing local overheating features from being overwhelmed by the global average; the fifth trend factor fitting unit uses least squares linear fitting to obtain the contact resistance change trend factor of each monitoring point, incorporating spatial distribution features and global statistics into the final thermal state feature vector, providing accurate, multi-dimensional, and distinguishable input for the anomaly discrimination module, significantly reducing false alarms and missed alarms, achieving accurate identification of complex operating conditions such as poor contact, overload, harmonic interference, and environmental fluctuations, supporting no-code adaptive configuration, and improving the operational safety of the power metering box and the reliability of the distribution network.

[0102] The anomaly detection module is configured to: compare the thermal state feature vector with a dynamic threshold set; and when it is detected that the absolute temperature value exceeds the first temperature threshold and the duration for which the temperature rise rate exceeds the second temperature rise rate threshold reaches the duration threshold, output an abnormal temperature rise confirmation signal; the dynamic threshold set includes a first temperature threshold, a second temperature rise rate threshold, and a duration threshold, and the dynamic threshold set is adaptively updated based on historical temperature statistics, ambient temperature compensation coefficient, and load current correction factor;

[0103] It should be further explained that the anomaly detection module in this embodiment includes:

[0104] It should be further explained that the first threshold adaptive unit in this embodiment is configured to: adaptively generate a dynamic threshold set for the current moment based on historical temperature statistics, ambient temperature compensation coefficient, and load current correction factor, and output the dynamic threshold set to the second condition determination unit; the specific implementation steps include:

[0105] D1 receives historical temperature statistics, current ambient temperature compensation coefficient, and current load current correction factor. The historical temperature statistics include a sequence of absolute temperature values ​​at each monitoring point over several past sampling periods and their corresponding temperature rise rate sequences. The ambient temperature compensation coefficient is equal to the deviation between the current ambient temperature and the standard ambient temperature divided by the standard ambient temperature. The load current correction factor is equal to the ratio of the current effective load current to the rated current. It should be further noted that in this embodiment, the standard ambient temperature is preset according to the rated operating ambient temperature range of the power metering box, for example, 25℃ or 40℃.

[0106] D2, based on the historical temperature statistics, calculate the upper limit of the absolute temperature value of each monitoring point under normal operating conditions, as the benchmark value of the first temperature threshold; calculate the upper limit of the temperature rise rate under normal operating conditions, as the benchmark value of the second temperature rise rate threshold; calculate the maximum duration for which the absolute temperature value under normal operating conditions exceeds the benchmark value of the first temperature threshold, as the benchmark value of the duration threshold; it should be further noted that, in this embodiment, the benchmark value of the duration threshold is the maximum duration of each instance where the temperature continuously exceeds the benchmark value of the first temperature threshold under normal operating conditions, which reflects the maximum heat accumulation time allowed by normal fluctuations.

[0107] D3. The reference value of the first temperature threshold is corrected according to the ambient temperature compensation coefficient. Specifically, when the ambient temperature compensation coefficient is greater than zero, it indicates that the current ambient temperature is higher than the standard ambient temperature, and the first temperature threshold is increased by a preset first proportional coefficient; when the ambient temperature compensation coefficient is less than zero, it indicates that the current ambient temperature is lower than the standard ambient temperature, and the first temperature threshold is decreased by a preset first proportional coefficient. It should be further noted that in this embodiment, the first proportional coefficient is determined based on experimental data on the thermal aging characteristics of the terminal material and the influence of ambient temperature on contact resistance. In actual scenarios, it is specifically set by those skilled in the art. For example, it is corrected by 0.5% to 2% for every 1°C deviation from the standard ambient temperature.

[0108] D4. The reference value of the second temperature rise rate threshold is corrected according to the load current correction factor. Specifically, when the load current correction factor is greater than one, it indicates that the current load current is greater than the rated current, and the second temperature rise rate threshold is increased by a preset second proportional coefficient. When the load current correction factor is less than one, it indicates that the current load current is less than the rated current, and the second temperature rise rate threshold is decreased by a preset second proportional coefficient. It should be further noted that in this embodiment, the second proportional coefficient is set according to the nonlinear influence curve of the load current on the terminal temperature rise rate. In actual scenarios, it is specifically set by those skilled in the art. For example, the threshold is increased by 3% to 5% for every 10% exceeding the rated current, and decreased by the same proportion when it is lower than the rated current. The determination of the normal operating state is based on the data segment after removing the marked abnormal events from the historical temperature statistics data. The 95th percentile of its absolute temperature value sequence is taken as the upper limit, and the 95th percentile of the temperature rise rate sequence is taken as the rate upper limit. D5. The corrected first temperature threshold, the corrected second temperature rise rate threshold, and the duration threshold calculated in the third step are combined together to form the dynamic threshold set at the current moment, and output to the second condition determination unit.

[0109] It should be further explained that the first threshold adaptive unit in this embodiment is further configured with upper and lower limit constraint logic for threshold correction, duration threshold setting rules under no over-temperature historical data, and validity verification rules for normal operation data segments; wherein, for the correction process of the first temperature threshold, the maximum value after correction is configured not to exceed 80% of the maximum allowable operating temperature of the insulation material corresponding to the terminal block, and the minimum value after correction is not lower than 50% of the first temperature threshold reference value under standard ambient temperature; for the correction process of the second temperature rise rate threshold, the maximum value after correction is configured not to exceed 3 times the upper limit of the normal temperature rise rate, and the minimum value after correction is not lower than 30% of the second temperature rise rate threshold reference value under standard rated current, thereby limiting the boundary range of threshold correction and avoiding abnormal misjudgment or fault omission caused by excessive threshold correction; for normal operation history In cases where there are no extreme scenarios where the absolute temperature value exceeds the first temperature threshold benchmark value in the data, the benchmark value for configuring the duration threshold is preferably 5 to 10 sampling periods. This benchmark value can be adaptively adjusted according to the sampling period length to effectively distinguish between instantaneous temperature fluctuations and continuous abnormal temperature rises. For the normal operation status data segment used for threshold benchmark value calculation, its validity verification rules are configured, that is, the cumulative duration of the data segment must be not less than 24 hours and the number of valid sampling points must be not less than 1000, so as to ensure the sample representativeness of the statistical data and avoid the distortion of the threshold benchmark value calculation caused by insufficient sample data. The specific parameter configuration of the above constraint logic, setting rules and verification rules in this embodiment can be adaptively adjusted by those skilled in the art according to the insulation temperature resistance level of the wiring terminals, equipment sampling parameters and on-site operating conditions, without departing from the protection scope of this invention.

[0110] It should be further explained that the second condition determination unit in this embodiment is configured as follows:

[0111] Based on the final thermal state feature vector and the dynamic threshold set, a multi-condition joint judgment is performed. When the judgment result meets the abnormal temperature rise condition, the monitoring point identifier that meets the condition and the corresponding thermal state feature vector are output to the third confirmation signal generation unit. Specifically:

[0112] E1. Extract the absolute temperature value, temperature rise rate, and duration of each monitoring point from the final thermal state feature vector; wherein, the duration refers to the cumulative length of time during which the absolute temperature value of the monitoring point continuously exceeds the first temperature threshold, and the temperature rise rate simultaneously exceeds the second temperature rise rate threshold during the cumulative time.

[0113] E2. For each monitoring point, determine whether the following three conditions are met simultaneously:

[0114] Condition 1: The absolute value of the current temperature is greater than the first temperature threshold.

[0115] Condition 2: The current temperature rise rate is greater than the second temperature rise rate threshold;

[0116] Condition 3: The current duration is greater than or equal to the duration threshold.

[0117] E3. If the current monitoring point meets all three judgment conditions at the same time, it is determined that the monitoring point meets the abnormal temperature rise condition, and the identifier of the monitoring point and the thermal state feature vector corresponding to the current time are output to the third confirmation signal generation unit; if it does not meet the condition, the monitoring data of the next sampling time are judged.

[0118] The third confirmation signal generation unit is configured to generate an abnormal temperature rise confirmation signal based on the monitoring point identifier that meets the abnormal temperature rise condition and the corresponding thermal state feature vector; wherein the abnormal temperature rise confirmation signal includes at least the following information: the monitoring point identifier where the abnormality occurred, the time when the abnormality occurred, the absolute temperature value when the abnormality occurred, the temperature rise rate, and the duration.

[0119] The abnormal temperature rise confirmation signal is output to the execution module.

[0120] The first threshold adaptive unit in this embodiment dynamically and adaptively generates the first temperature threshold, the second temperature rise rate threshold, and the duration threshold by introducing historical temperature statistics, ambient temperature compensation coefficient, and load current correction factor. It further configures upper and lower limit constraint logic for threshold correction, duration threshold setting rules under conditions of no over-temperature historical data, and validity verification rules for normal operation data segments. This effectively avoids threshold inaccuracies caused by ambient temperature fluctuations, load current changes, and insufficient sample data, significantly improving the environmental adaptability and robustness of anomaly detection. Simultaneously, the second condition determination unit uses a logic that jointly determines the absolute temperature value, temperature rise rate, and duration, combined with cumulative monitoring of duration, to accurately distinguish between instantaneous temperature fluctuations and continuous abnormal temperature rises, greatly reducing the false alarm rate caused by load mutations or electromagnetic interference. The third confirmation signal generation unit outputs a complete confirmation signal containing monitoring point identification, abnormal time, and characteristic parameters, providing reliable data support for subsequent hierarchical early warning and fault tracing in the execution module. In summary, this embodiment achieves accurate identification and reliable confirmation of abnormal temperature rise of wiring terminals through dynamic threshold adaptation, multi-condition joint judgment, and integrity signal output, thereby improving the operational safety and intelligent operation and maintenance level of the power metering box.

[0121] Example 2

[0122] Please see Figure 2 However, further explanation is needed. The execution module in this embodiment is configured to: execute an abnormal temperature rise confirmation signal, including:

[0123] When the abnormal temperature rise confirmation signal is received, the warning output module is driven to output primary warning information in a preset manner and acquire online monitoring data under abnormal conditions. The online monitoring data includes the real-time value of the contact resistance change trend factor and the fluctuation rate of the current load current. In this embodiment, the configuration of the preset manner refers to the preset combination rules of the output channel, information format, and push strategy of the primary warning information when the warning output module receives the abnormal temperature rise confirmation signal. Specifically, the output channel can be configured according to the operation and maintenance management requirements as one or more parallel methods among local sound and light alarms, operation and maintenance platform pop-ups, SMS push, and mobile operation and maintenance APP messages. The information format is at least... It includes the abnormal monitoring point number, abnormal occurrence timestamp, real-time temperature value, temperature rise rate, and corresponding circuit load current value. The push strategy can be set with different sending frequencies and receiving object groups according to the abnormality level or time period (e.g., level 1 abnormalities are only pushed to the area inspectors during daytime, while level 2 and above abnormalities are pushed to the shift leader and maintenance team in real time all day). In this embodiment, all parameters of the preset mode can be graphically set through configuration files or maintenance platform interface during the system initialization stage, and online dynamic modification is supported. It can be routinely adjusted by those skilled in the art according to the installation environment of the metering box, the maintenance organization structure, and the early warning response time requirements, without departing from the protection scope of this invention.

[0124] If the contact resistance change trend factor is less than the third preset threshold and the fluctuation rate of the current load current is less than the fourth preset threshold, it is determined to be a real abnormal temperature rise. The synchronous control storage module records the current abnormal event and outputs a warning completion command. In this embodiment, the third preset threshold is a boundary value used to determine whether the contact resistance change trend is within the normal range. Its configuration is based on the historical operating data statistics of the terminal block under normal contact conditions: the contact resistance change trend factor sequence calculated by the fifth trend factor fitting unit is obtained for each monitoring point in multiple complete load cycles in which there is no known abnormal temperature rise and the contact state has been manually confirmed to be good. The 95th percentile or the upper limit of three standard deviations of this sequence is used as the benchmark value of the third preset threshold. Since the temperature change and the effective value change of the fundamental current maintain a relatively stable linear proportional relationship under normal operating conditions, the contact resistance change trend factor fluctuates only within a small range. This benchmark value usually tends to a small positive value, and the typical value can be set between 0.05 and 0.15. If the calculated contact resistance change trend factor is less than the threshold, it indicates that the contact resistance has not shown a significant nonlinear increasing trend, and the temperature rise is mainly caused by normal heating from a continuous high current. Combined with the load current fluctuation rate, this can be confirmed as a genuine abnormal temperature rise. Conversely, if it is greater than or equal to the threshold, it indicates that the ratio of temperature change to current change has deviated from the normal stable state, possibly caused by unstable contact resistance due to oxidation or loosening of the contact surface. A retry counter needs to be activated for further verification. The specific value of the aforementioned third preset threshold can be routinely calibrated and adjusted by those skilled in the art based on the contact material characteristics of the terminals, the rated current level, and the accumulated historical operating data, all without departing from the scope of protection of this invention.

[0125] If the contact resistance change trend factor is less than the third preset threshold, but the fluctuation rate of the current load current is greater than or equal to the fourth preset threshold, it is determined to be an instantaneous temperature response caused by a sudden load change. The abnormal event is not recorded temporarily, the current temperature sequence is maintained, and the abnormality detection module is re-executed after a one-sampling-cycle delay. It should be further noted that this embodiment, when re-executing the abnormality detection module after a one-sampling-cycle delay, also includes a conflict handling process during the delay period, specifically:

[0126] If an abnormal temperature rise confirmation signal is received again from the same monitoring point during the delay period, the signal is ignored and the judgment is not repeated.

[0127] After the delay ends, when re-entering the anomaly detection module, the temporary status flags accumulated during the delay period are cleared, and the latest temperature sequence and load current data are used as input.

[0128] If the number of consecutive ignored abnormal temperature rise confirmation signals exceeds the seventh preset threshold during the delay period, it will be directly determined as a real abnormal temperature rise, and the remaining discrimination will be skipped to directly output an early warning.

[0129] If the contact resistance change trend factor is greater than or equal to the third preset threshold, the retry counter is started. When the retry count is less than the fifth preset threshold, the current temperature sequence is cleared and the thermal state feature extraction step is returned to be executed. At the same time, the retry count is incremented by one. When the retry count reaches the fifth preset threshold, it is determined to be a continuous unstable contact state. A warning completion command is forcibly output and the retry counter is reset.

[0130] It should be further explained that, in this embodiment, the specific implementation of the abnormality level classification step is as follows: when it is determined to be a real abnormal temperature rise, the difference between the current absolute temperature value and the first temperature threshold is obtained, and the temperature exceedance ratio is calculated. The temperature exceedance ratio is equal to the difference divided by the first temperature threshold.

[0131] In this embodiment, when executing the abnormal temperature rise confirmation signal, an abnormality level classification step is also included, specifically: when the temperature exceedance ratio is less than the first level threshold, the abnormality level is marked as a level one abnormality, and the warning output module is controlled to output a yellow warning message; when the temperature exceedance ratio is between the first level threshold and the second level threshold, the abnormality level is marked as a level two abnormality, an orange warning message is output, and on-site inspection is recommended; when the temperature exceedance ratio is greater than or equal to the second level threshold, the abnormality level is marked as a level three abnormality, a red warning message is output, and a remote power-off protection command is triggered. As an exemplary method of value determination, in this embodiment, the configuration of the first and second level thresholds is based on the temperature resistance rating of the terminal insulation material, thermal aging characteristic test data, and the graded handling strategy for the severity of abnormalities in operation and maintenance management: the first level threshold (10%) corresponds to a slight overheating state, and its setting is based on the overheating starting point of the insulation material with a safety margin on the basis of the long-term allowable operating temperature. It is usually taken as 5% to 15% of the allowable temperature rise value of the insulation material. When the temperature over-limit ratio is lower than this value, the terminal is still within the safe boundary of the material's thermal resistance, but there is a trend of deviating from normal operation, which needs to be noted; the second level threshold (30%) corresponds to a slight overheating state. The percentage (%) corresponds to the critical point of severe overheating. It is set based on the accelerated deterioration inflection point in the thermal aging life curve of the insulation material. That is, after the overheating exceeds this percentage, the insulation life will decrease at an exponential rate. At the same time, the combined risks of accelerated oxidation of metal conductors and increased fretting wear of contact surfaces are considered. The typical value can refer to the upper limit of the overheating percentage (30%~40%) of the "severe abnormality" level in the electrical equipment temperature rise test standard. In this embodiment, the specific values ​​of the first level threshold and the second level threshold can be routinely adjusted by those skilled in the art according to the actual insulation material model of the terminal block, the rated operating temperature level and the operation and maintenance procedures of the power distribution network, without departing from the protection scope of this invention.

[0132] It should be further noted that this embodiment also includes a retry historical accumulation recording step when executing the abnormal temperature rise confirmation signal, specifically as follows:

[0133] An independent cumulative instability counter is maintained for each monitoring point. This counter is used to record the number of times the monitoring point reaches the fifth preset threshold and triggers a mandatory warning within a preset time window.

[0134] When the retry count reaches the fifth preset threshold and a warning completion command is forcibly output, the accumulated instability counter is incremented by one, and the current timestamp is recorded.

[0135] When the accumulated instability counter exceeds the sixth preset threshold, it is determined that the monitoring point has a chronic contact degradation trend, and a chronic degradation warning message is generated and pushed to the operation and maintenance platform.

[0136] Each monitoring point is maintained independently with an accumulated instability counter, which records the cumulative number of times that monitoring point triggers a mandatory early warning.

[0137] When the retry count reaches the fifth preset threshold and a warning completion command is forcibly output, the accumulated instability counter is incremented by one, and the current timestamp is recorded.

[0138] When the accumulated instability counter exceeds the sixth preset threshold, it is determined that the monitoring point has a chronic contact degradation trend, and a chronic degradation warning message is generated and pushed to the operation and maintenance platform.

[0139] Every preset reset period, if the monitoring point does not trigger a mandatory warning within that period, the accumulated instability counter will be decremented by one or cleared to zero.

[0140] It should be further explained that in the retry history accumulation recording step of this embodiment, the preset values ​​are set according to the following: The fifth preset threshold represents the maximum number of consecutive retries allowed, which is determined based on the thermal response time constant and sampling period of the terminal under normal contact conditions, and is usually set to 3 to 5 times to exclude transient unstable states caused by instantaneous electromagnetic interference or load spikes; the sixth preset threshold represents the upper limit of the number of times a mandatory warning is triggered within a preset time window, which is set based on the cumulative thermal fatigue characteristics of the terminal material and on-site operation and maintenance experience, and is usually set to 3 times. Exceeding this value indicates that the contact resistance has a continuous deterioration trend; the preset time window is determined based on the terminal heat dissipation and thermal recovery cycle, and is usually set to 24 hours or 72 hours to cover a complete daily load fluctuation cycle; the preset reset cycle is set based on the equipment's regular inspection interval or seasonal load change cycle, and is usually set to seven days or thirty days. If no mandatory warning is triggered again, the accumulated count is gradually reduced to reflect the self-recovery characteristics of the contact state. All of the above thresholds can be calibrated through historical operating data or accelerated aging tests before leaving the factory.

[0141] It should be further explained that the execution module of this embodiment is further configured with the reset trigger logic of the retry counter, the output specification of the primary warning information, the interlock execution logic of the remote power failure protection command, and the reporting mechanism of the hierarchical warning information; among them, for the retry counter, its reset trigger logic is configured as follows: when the single thermal state feature extraction process in the retry process is completed, if the anomaly discrimination module determines that no abnormal temperature rise condition is met, the retry counter of the corresponding monitoring point is immediately cleared to zero. After the equipment completes the power failure and restart operation, the retry counters corresponding to all monitoring points are reset to the initial value of 0; for The primary warning information is configured to include the following output specifications: a unique identifier for the abnormal monitoring point, a timestamp of the abnormality, the real-time absolute temperature value of the abnormal monitoring point, the real-time temperature rise rate, and the real-time load current value of the corresponding circuit. The primary warning information is output via at least one of the following methods: local audible and visual warning unit, pop-up window on the maintenance platform, SMS push notification from maintenance personnel, or message push notification from the mobile maintenance APP. For remote power outage protection commands, the interlocking execution logic is configured as follows: when any of the following interlocking trigger conditions are detected, the remote power outage protection command for the corresponding circuit is interlocked, and only a warning signal is output. The following conditions will prevent a power outage from being executed: firstly, if an equipment maintenance interlock command is received from the operation and maintenance platform; secondly, if the current load current of the corresponding circuit exceeds 1.2 times its rated current, and a power outage would pose a risk of equipment damage or a safety accident; and thirdly, if the corresponding power supply circuit is detected as a primary load critical circuit such as fire protection or emergency lighting. Simultaneously, the interlock status command and the corresponding interlock trigger reason will be output to the operation and maintenance platform along with the warning information. For tiered warning information, the reporting mechanism is configured as follows: Level 1 abnormal warning information will be reported to the operation and maintenance platform at a frequency of once per hour; Level 2 abnormal warning information will be reported at a frequency of once every 15 minutes. The rate is reported to the operation and maintenance platform. The three-level abnormality warning information is reported in real time until the corresponding abnormal state is resolved. After the abnormal state is resolved, the corresponding abnormal event processing report is automatically generated. The report includes at least the total duration of the abnormality, the peak data during the abnormal process, and the abnormality handling result. After the processing report is completed, it is archived and stored in the storage module. The specific parameter configuration of the above reset logic, output specification, interlocking logic and reporting mechanism in this embodiment can be adapted by those skilled in the art according to the metering box circuit level, on-site operation and maintenance management requirements, and equipment rated electrical parameters, without departing from the protection scope of this invention.

[0142] The execution module in this embodiment introduces a dual criterion of contact resistance change trend factor and load current fluctuation rate, combined with a retry counter and delay conflict handling mechanism, to accurately distinguish between real abnormal temperature rise, instantaneous response to load changes, and continuous unstable contact states. This effectively eliminates false alarms caused by non-fault factors such as electromagnetic interference and load spikes, significantly reducing the false alarm rate. Simultaneously, the anomaly level classification step divides anomalies into three levels based on the temperature exceedance ratio, outputting yellow, orange, and red warnings and remote power-off protection commands respectively. This allows maintenance personnel to quickly determine the response priority based on the warning level. This system avoids blindly shutting down power due to minor overheating or delaying handling due to severe overheating. The remote power outage protection command's interlocking execution logic further considers maintenance status, overload risk, and critical load protection needs, preventing secondary equipment damage or safety accidents caused by improper power outages. The tiered early warning reporting mechanism dynamically adjusts the reporting frequency according to the anomaly level, ensuring real-time response to important anomalies while avoiding the impact of frequent reporting of primary anomalies on the operation and maintenance platform. The retry history accumulation recording step effectively identifies chronic contact degradation trends through an accumulated instability counter and a time window sliding decay mechanism, providing data support for preventive maintenance. In summary, this embodiment achieves intelligent execution across the entire chain, from anomaly identification, tiered early warning, safe power outage to trend prediction, significantly improving the reliability, safety, and operation and maintenance efficiency of temperature rise management of the power metering box terminals.

[0143] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the claims. All of these variations are within the protection scope of the present invention.

Claims

1. An automatic inspection system for temperature rise of wiring terminals in an electricity metering box, characterized in that, include: The data acquisition module is configured to acquire real-time temperature sequences, ambient temperature data, and load current data of the terminal blocks. The feature extraction module is configured to: perform thermal state feature extraction based on the real-time temperature sequence, ambient temperature data, and load current data, and generate a thermal state feature vector. The thermal state feature vector includes absolute temperature value, temperature rise rate, duration, and contact resistance change trend factor. The temperature rise rate is calculated from the temperature difference and time interval between adjacent sampling times, and the contact resistance change trend factor is obtained by fitting the ratio of temperature change to load current change. The anomaly detection module is configured to: compare the thermal state feature vector with a dynamic threshold set; and when it is detected that the absolute temperature value exceeds the first temperature threshold in the dynamic threshold set, and the duration for which the temperature rise rate exceeds the second temperature rise rate threshold in the dynamic threshold set reaches the duration threshold in the dynamic threshold set, output an abnormal temperature rise confirmation signal. The execution module is configured to: execute an abnormal temperature rise confirmation signal.

2. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 1, characterized in that, The execution of the abnormal temperature rise confirmation signal includes: When the abnormal temperature rise confirmation signal is received, the driver outputs the early warning information in a preset manner and obtains the online monitoring data under the abnormal state. The online monitoring data includes the real-time value of the contact resistance change trend factor and the fluctuation rate of the current load current. If the contact resistance change trend factor is less than the third preset threshold and the fluctuation rate of the current load current is less than the fourth preset threshold, it is determined to be a real abnormal temperature rise. The synchronous control storage module records the current abnormal event and outputs a warning completion command.

3. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 2, characterized in that, The execution of the abnormal temperature rise confirmation signal also includes: If the contact resistance change trend factor is less than the third preset threshold, but the fluctuation rate of the current load current is greater than or equal to the fourth preset threshold, it is determined to be an instantaneous temperature response caused by a sudden load change. The abnormal event is not recorded for the time being. The current temperature sequence is maintained and the abnormal judgment module is re-executed after a delay of one sampling period. If the contact resistance change trend factor is greater than or equal to the third preset threshold, the retry counter is started. When the retry count is less than the fifth preset threshold, the current temperature sequence is cleared and the thermal state feature extraction step is returned to be executed. At the same time, the retry count is incremented by one. When the retry count reaches the fifth preset threshold, it is determined to be a continuous unstable contact state. A warning completion command is forcibly output and the retry counter is reset.

4. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 3, characterized in that, The feature extraction module includes: The first preprocessing unit is configured to perform adaptive filtering on the real-time temperature sequence of the terminals to obtain a smooth temperature sequence; wherein the adaptive filtering adopts a series structure of sliding window mid-range filtering and first-order hysteresis filtering; the window length L of the sliding window mid-range filtering is based on the load current fluctuation rate σ. I The adjustments are as follows: σ I It equals the ratio of the standard deviation to the mean of the load current sequence within the current sliding window; when σ I Greater than the preset fluctuation threshold δ I When σ is reached, L is set as the first length L1; when σ is reached... I Less than or equal to δ I When L is set to the second length L2, and L1 > L2; The filter coefficients of the first-order hysteresis filter It is calculated adaptively based on the local variance of the temperature series, specifically as follows: ,in For the estimated noise variance, This represents the variance of the local signal change over k temperature values ​​before and after the current sampling point.

5. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 4, characterized in that, The feature extraction module further includes: The second phase compensation unit is configured to: receive the smoothed temperature sequence and load current data, perform phase compensation on the load current data, and obtain an equivalent excitation current sequence; the equivalent excitation current sequence and the smoothed temperature sequence are aligned in phase, such that the absolute value of the phase difference between the theoretical response temperature after the equivalent excitation current is subjected to the terminal thermal inertia transfer function and the smoothed temperature sequence is less than a preset phase difference threshold. The third fundamental frequency separation unit is configured to: receive the equivalent excitation current sequence, perform fundamental frequency separation processing on the equivalent excitation current sequence to obtain the effective value sequence of the fundamental frequency current, and filter out the harmonic components in the equivalent excitation current sequence.

6. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 5, characterized in that, The anomaly detection module includes a first threshold adaptive unit, configured to generate a dynamic threshold set for the current moment based on historical temperature statistics, ambient temperature compensation coefficient, and load current correction factor; the dynamic threshold set includes a first temperature threshold, a second temperature rise rate threshold, and a duration threshold.

7. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 6, characterized in that, The anomaly detection module further includes: a second condition determination unit, configured to: perform multi-condition joint determination based on the final thermal state feature vector and the dynamic threshold set; when the determination result meets the abnormal temperature rise condition, output the monitoring point identifier that meets the condition and the corresponding thermal state feature vector to the third confirmation signal generation unit, specifically: E1. Extract the absolute temperature value, temperature rise rate, and duration of each monitoring point from the final thermal state feature vector; wherein, the duration refers to the cumulative length of time during which the absolute temperature value of the monitoring point continuously exceeds the first temperature threshold, and the temperature rise rate simultaneously exceeds the second temperature rise rate threshold during the cumulative time. E2. For each monitoring point, determine whether the following three conditions are met simultaneously: Condition 1: The absolute value of the current temperature is greater than the first temperature threshold. Condition 2: The current temperature rise rate is greater than the second temperature rise rate threshold; Condition 3: The current duration is greater than or equal to the duration threshold. E3. If the current monitoring point meets all three judgment conditions at the same time, it is determined that the monitoring point meets the abnormal temperature rise condition, and the identifier of the monitoring point and the thermal state feature vector corresponding to the current time are output to the third confirmation signal generation unit; if it does not meet the condition, the monitoring data of the next sampling time are judged.

8. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 7, characterized in that, The execution of the abnormal temperature rise confirmation signal also includes an abnormality level classification step, specifically: When a true abnormal temperature rise is determined, the difference between the absolute value of the current temperature and the first temperature threshold is obtained, and the temperature exceedance ratio is calculated. The temperature exceedance ratio is equal to the difference divided by the first temperature threshold. When the temperature exceedance ratio is less than the first-level threshold, the abnormality level is marked as a first-level abnormality, and the warning output module is controlled to output a yellow warning message. When the temperature exceedance ratio is between the first-level threshold and the second-level threshold, the abnormality level is marked as a level two abnormality, an orange warning message is output, and on-site inspection is recommended. When the temperature exceedance ratio is greater than or equal to the second-level threshold, the abnormality level is marked as a third-level abnormality, a red warning message is output, and a remote power-off protection command is triggered.

9. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 8, characterized in that, The execution of the abnormal temperature rise confirmation signal also includes a retry historical accumulation record step, specifically: An independent cumulative instability counter is maintained for each monitoring point. This counter is used to record the number of times the monitoring point reaches the fifth preset threshold and triggers a mandatory warning within a preset time window. When the retry count reaches the fifth preset threshold and a warning completion command is forcibly output, the accumulated instability counter is incremented by one, and the current timestamp is recorded. When the accumulated instability counter exceeds the sixth preset threshold, it is determined that the monitoring point has a chronic contact degradation trend, and a chronic degradation early warning message is generated and pushed to the operation and maintenance platform.

10. The automatic temperature rise inspection system for the wiring terminals of an energy metering box as described in claim 9, characterized in that, When the anomaly detection module is re-executed after a one-sampling-cycle delay, a conflict handling process during the delay period is also included, specifically: If an abnormal temperature rise confirmation signal is received again from the same monitoring point during the delay period, the signal is ignored and the judgment is not repeated. After the delay ends, when re-entering the anomaly detection module, the temporary status flags accumulated during the delay period are cleared, and the latest temperature sequence and load current data are used as input. If the number of consecutive ignored abnormal temperature rise confirmation signals exceeds the seventh preset threshold during the delay period, it will be directly determined as a real abnormal temperature rise, and the remaining discrimination will be skipped to directly output an early warning.

Images