A method and system for monitoring faults of electric meter box wire clamps

Abstract

The present application provides a method and system for monitoring the fault of a meter box wire clamp, which relates to the field of power equipment monitoring technology. The key points of its technical solution are: obtaining the temperature measurement value of a single position in the meter box, the current measurement values of phase A, phase B, and phase C of the meter box incoming line, and the neutral line current measurement value; calculating the expected temperature value in the box; calculating the temperature deviation value between the temperature measurement value and the expected temperature value in the box; calculating the heating index value related to each current; calculating and tracking the dynamic correlation index between the time series of the temperature deviation value and the time series of the respective heating index values; judging whether the wire clamp of the specific line has experienced early degradation based on the changing trend of the dynamic correlation index. The method and system for monitoring the fault of a meter box wire clamp provided by the present application have the advantages of low cost, the ability to effectively identify and distinguish which phase or neutral line wire clamp has shown signs of early degradation, strong anti-interference ability, and sensitivity to early fault signals.

Description

Technical Field

[0001] The present application relates to the technical field of power equipment monitoring, and in particular to a method and system for monitoring faults in an electric meter box wire clamp. Background Art

[0002] In low-voltage power distribution systems in urban residential areas or commercial buildings, the meter box serves as a critical interface between the power supply network and end users, and its operating status directly impacts power supply reliability. In these scenarios, user-side electrical devices are mostly single-phase, and their power consumption exhibits significant time-varying and random characteristics. This characteristic results in the low-voltage distribution network to which the meter box is connected experiencing a long-term three-phase load imbalance. This means that the effective current flowing through the A, B, and C phases at different times varies significantly, with the current in one phase potentially exceeding that of the other two. According to circuit principles, this three-phase unbalanced current generates a current on the neutral conductor (N conductor). Especially in the presence of a large number of nonlinear loads, the resulting third harmonic current is superimposed on the N conductor, potentially causing the N conductor current to approach or even exceed the phase current.

[0003] Inside the meter box, the incoming power cables connect through the corresponding main clamps for phases A, B, C, and N. The current is then distributed to single-phase or three-phase energy meters via the clamps at the front of each user's meter. This means that all clamps for a particular phase (such as phase A) (the main clamp and all user clamps at the front of that phase) jointly carry the total current and the distributed current for that phase. Due to the long-term imbalance in the three-phase load, the electrical load levels experienced by the clamps of each phase vary continuously. For example, the clamp group corresponding to the chronically heavily loaded phase A will experience higher average currents and more severe current fluctuations, resulting in greater electrothermal stress cycles. In contrast, the clamps of the lightly loaded phase C experience less electrothermal stress. This differentiated operating condition based on phase directly leads to different degradation rates and patterns for the clamps of each phase.

[0004] The reliability of a clamp connection depends on the physical and chemical state of the contact surface and the clamping force. For clamps placed on phases subject to long-term heavy loads (such as phase A), the greater thermal expansion and contraction can accelerate the evolution of the contact surface's microstructure, the formation and destruction of oxide layers, and the reduction in preload in fasteners (such as bolts) due to material creep. Consequently, early signs of degradation, such as a slow increase in contact resistance or unstable contact, may appear sooner than for clamps placed on phases subject to light loads (such as phase C). Furthermore, the N-line clamp carries not only the fundamental current generated by the imbalance but also the zero-sequence harmonic currents generated by the nonlinear loads of each phase. These high-frequency components can induce additional mechanical vibration and non-uniform heating, resulting in unique degradation effects.

[0005] However, this differential early degradation between the clamps of different phase lines (and the neutral line) caused by three-phase load imbalance often manifests itself in subtle physical manifestations (such as abnormal temperature rise) in the initial stages of fault development. This is easily obscured or masked by natural temperature fluctuations within the meter box and the heat diffusion effects of other heat-generating components within the box (such as energy meters and circuit breakers). Traditional monitoring methods, such as regular infrared temperature measurement, can miss early faults due to limited time windows or weak signals. Furthermore, deploying high-precision, specialized sensors (such as temperature and vibration sensors) for each of the numerous clamps faces significant obstacles in terms of cost, installation space, and data processing complexity, making large-scale deployment unfeasible.

[0006] In view of the above problems, the existing technology is in urgent need of improvement. Summary of the Invention

[0007] The purpose of this application is to provide a meter box wire clamp fault monitoring method and system, which has the advantages of low cost, effective identification and distinction of which phase or neutral wire clamp shows early signs of degradation, strong anti-interference ability, and sensitivity to early fault signals.

[0008] In the first aspect, the present application provides a method for monitoring faults in an electric meter box wire clamp, the technical solution of which is as follows:

[0009] include:

[0010] Obtaining a temperature measurement value of a single location in the electric meter box;

[0011] Obtaining the current measurement values of phase A, phase B, phase C and neutral line of the meter box incoming line;

[0012] Calculating an expected temperature value inside the box based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and a baseline thermal model representing the health status of the clamp;

[0013] Calculating a temperature deviation between the temperature measurement value and the expected temperature value in the box;

[0014] Based on the measured current values of the phases A, B, and C and the measured current value of the neutral line, respectively, calculating the current-related heating index values of the phase A line, the phase B line, the phase C line, and the neutral line;

[0015] Calculate and track a dynamic correlation index between a time series of the temperature deviation value and a time series of heating index values of each of the A-phase line, the B-phase line, the C-phase line, and the N-line line;

[0016] According to the changing trend of the dynamic correlation index, when the dynamic correlation index corresponding to a specific line shows an enhancement and reaches a preset judgment condition, it is determined that the wire clamp of the specific line has undergone early degradation;

[0017] When the electric meter box operates in a low-load cycle, the parameters in the reference thermal model are calibrated using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values and the neutral current measurement value obtained in the cycle.

[0018] Furthermore, in the present application, the step of calculating and tracking the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line includes:

[0019] Calculate the time series of the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index value of each line;

[0020] Set the discrimination threshold of the correlation index;

[0021] Set the duration threshold;

[0022] For the time series of the dynamic correlation index of a specific line, determining whether its value continuously exceeds the discrimination threshold of the correlation index within a time length defined by the discrimination threshold of the duration;

[0023] When the time series of the dynamic correlation index of the specific line meets the continuous exceeding condition, the enhancement trend presented by the dynamic correlation index is recognized as a real enhancement caused by continuous degradation.

[0024] Furthermore, in the present application, when the electric meter box operates in a low-load cycle, the step of calibrating the parameters in the reference thermal model using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values, and the neutral current measurement value obtained during the cycle includes:

[0025] Setting a first threshold value for defining a current amplitude of a low load;

[0026] setting a second threshold for defining a current variation indicator of load stability and a length of a time window based on which the current variation indicator is calculated;

[0027] Obtaining the current measurement values of phase A, phase B, phase C and neutral line of the meter box incoming line;

[0028] determining whether the current measurement value is lower than the first threshold;

[0029] Calculating the current change index within the time window based on the current measurement value and the time window length;

[0030] Determining whether the current change index is lower than the second threshold;

[0031] When the current measurement value is lower than the first threshold and the current change index is lower than the second threshold, determining the current operation cycle as the low-load cycle;

[0032] During the determined low-load period, the parameters in the reference thermal model are calibrated using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values and the neutral line current measurement value obtained during the period.

[0033] Furthermore, in the present application, the step of calculating a dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line includes:

[0034] Obtain the quantitative value of historical fluctuation characteristics of line load;

[0035] Obtain information about the identified degradation stage of the line;

[0036] Determining a time window length for calculating the dynamic correlation index based on the historical fluctuation characteristic quantified value and the identified degradation stage information;

[0037] The determined time window length is used to calculate a dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line.

[0038] Furthermore, in the present application, the step of determining the length of the time window for calculating the dynamic correlation index based on the historical fluctuation characteristic quantization value and the identified degradation stage information includes:

[0039] defining discrete levels of the quantized value of the historical fluctuation characteristic and discrete levels of the identified degradation stage information;

[0040] Constructing a two-dimensional lookup table, using the discrete levels of the historical fluctuation characteristic quantization value and the discrete levels of the identified degradation stage information as indexes, and presetting a reference time window length value in each cell of the two-dimensional lookup table;

[0041] Obtaining the reference time window length value from the two-dimensional lookup table according to the discrete level corresponding to the quantized value of the historical fluctuation characteristic and the discrete level corresponding to the identified degradation stage information;

[0042] Calculating the rate of change of the dynamic correlation index within a preset time period;

[0043] determining an adjustment factor according to the rate of change of the dynamic correlation index;

[0044] The obtained reference time window length value is calculated with the adjustment factor to obtain the time window length used to calculate the dynamic correlation index.

[0045] Furthermore, in the present application, the step of determining an adjustment factor according to the rate of change of the dynamic correlation index includes:

[0046] Presetting a plurality of numerical ranges of the dynamic correlation index change rates;

[0047] Presetting a corresponding adjustment factor value for each of the numerical intervals;

[0048] Comparing the calculated change rate of the dynamic correlation index with the multiple numerical intervals to determine the numerical interval to which the change rate of the dynamic correlation index belongs;

[0049] A preset adjustment factor value corresponding to the numerical range to which the rate of change of the dynamic correlation index belongs is selected as the determined adjustment factor.

[0050] Furthermore, in the present application, the step of calculating the rate of change of the dynamic correlation index within a preset time period includes:

[0051] Obtaining the dynamic correlation index value corresponding to the start time of the preset time period;

[0052] Obtaining the dynamic correlation index value corresponding to the end time of the preset time period;

[0053] Calculating the numerical difference between the dynamic correlation index value at the end time and the dynamic correlation index value at the start time;

[0054] The numerical difference is divided by the length of the preset time period to obtain the change rate of the dynamic correlation index within the preset time period.

[0055] Furthermore, in the present application, the step of calculating an expected box internal temperature value based on the A-phase, B-phase, and C-phase current measurements, the neutral line current measurement, and a baseline thermal model representing the health status of the clamp includes:

[0056] A physical model based on the heat circuit network principle is used as the reference thermal model;

[0057] Calculating the healthy heating power of each line based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and the preset contact resistance value of each line in a healthy state, and setting the healthy heating power as the heat input of the heat source node in the benchmark thermal model;

[0058] Setting thermal resistance nodes representing the heat transfer path inside the meter box and the heat dissipation path between the meter box and the external environment in the benchmark thermal model;

[0059] Setting nodes representing the thermal capacity of the electric meter box and its internal components in the baseline thermal model;

[0060] The benchmark thermal model including the heat source node, the thermal resistance node, and the heat capacity node is solved to obtain the expected temperature value in the box.

[0061] Furthermore, in the present application, the step of calculating and tracking the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line includes:

[0062] Setting the length of the sliding time window and the sliding step size for calculating the dynamic correlation index;

[0063] At each sliding time window position, obtaining a time series segment of the temperature deviation value within the sliding time window;

[0064] Obtaining a time series segment of a heating index value of any one of the A-phase line, the B-phase line, the C-phase line, or the N-line line corresponding to the position of the sliding time window;

[0065] Calculating a Pearson correlation coefficient between the acquired time series segment of the temperature deviation value and the acquired time series segment of the heating index value of the specific circuit;

[0066] Assigning the calculated Pearson correlation coefficient as a dynamic correlation index value of a specific line corresponding to the position of the sliding time window;

[0067] As the sliding time window moves according to the sliding step, the steps of obtaining the time series segment of the temperature deviation value, obtaining the time series segment of the heating index value of the specific line, calculating the Pearson correlation coefficient, and assigning the value to the dynamic correlation index are repeated to form the dynamic correlation index of each specific line.

[0068] In a second aspect, the present application also proposes a meter box wire clamp fault monitoring system, which includes:

[0069] A temperature acquisition module, used to obtain a temperature measurement value of a single location in the meter box;

[0070] A current acquisition module is used to obtain the current measurement values of phase A, phase B, phase C and neutral line of the meter box incoming line;

[0071] an expected temperature calculation module, configured to calculate an expected internal temperature value of the box based on the measured current values of the phases A, B, and C, the measured current value of the neutral line, and a reference thermal model representing the health status of the clamp;

[0072] a temperature deviation calculation module, configured to calculate a temperature deviation between the temperature measurement value and the expected temperature value inside the box;

[0073] a heating index calculation module, configured to calculate current-related heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line based on the A-phase, B-phase, and C-phase current measurement values and the neutral line current measurement value;

[0074] a correlation calculation and tracking module, configured to calculate and track a dynamic correlation index between a time series of the temperature deviation value and a time series of heating index values of each of the A-phase line, the B-phase line, the C-phase line, and the N-line line;

[0075] a fault determination module, configured to determine, based on a change trend of the dynamic correlation index, that the wire clamp of a specific line has experienced early degradation when the dynamic correlation index corresponding to the specific line shows an enhancement and reaches a preset judgment condition;

[0076] The model calibration module is used to calibrate the parameters in the reference thermal model when the meter box operates in a low load cycle by using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values and the neutral line current measurement value obtained in the cycle.

[0077] From the above, it can be seen that the present application provides a meter box wire clamp fault monitoring method and system, which can effectively identify and locate the early degradation faults of specific phase or neutral wire clamps in the meter box by utilizing a single temperature sensor and existing current measurement data, combined with a benchmark thermal model and dynamic correlation analysis. It has the advantages of low cost, the ability to effectively identify and distinguish which phase or neutral wire clamp has shown signs of early degradation, strong anti-interference ability, sensitivity to early fault signals, and adaptive model. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 This is a flow chart of a method for monitoring faults in an electric meter box wire clamp provided in this application.

[0079] Figure 2This is a structural diagram of an electric meter box wire clamp fault monitoring system provided in this application.

[0080] In the figure: 1. Temperature acquisition module; 2. Current acquisition module; 3. Expected temperature calculation module; 4. Temperature deviation calculation module; 5. Heat index calculation module; 6. Correlation calculation and tracking module; 7. Fault judgment module; 8. Model calibration module. DETAILED DESCRIPTION

[0081] The technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. The components of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for which protection is claimed, but merely represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of this application.

[0082] Reference Figure 1 , this application proposes a method for monitoring faults in an electric meter box wire clamp, comprising:

[0083] S110, obtaining a temperature measurement value of a single location in the meter box;

[0084] S120, obtaining the measured current values of phases A, B, and C of the incoming line of the meter box and the measured current value of the neutral line;

[0085] S130, calculating an expected temperature value inside the box based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and a baseline thermal model representing the health status of the wire clamp;

[0086] S140, calculating the temperature deviation between the temperature measurement value and the expected temperature value in the box;

[0087] S150, based on the measured current values of phase A, phase B, phase C and the measured current value of the neutral line, respectively calculating the current-related heating index values of the phase A line, the phase B line, the phase C line and the neutral line;

[0088] S160, calculating and tracking a dynamic correlation index between a time series of temperature deviation values and a time series of heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line;

[0089] S170: When, based on a change trend of the dynamic correlation index, the dynamic correlation index corresponding to a specific line shows an enhancement and reaches a preset judgment condition, it is determined that the wire clamp of the specific line has experienced early degradation;

[0090] S180 , when the electric meter box operates in a low load cycle, calibrate parameters in the reference thermal model using temperature measurement values and A-phase, B-phase, and C-phase current measurement values, and neutral line current measurement values obtained during the cycle.

[0091] Acquiring a temperature measurement at a single location within the meter box involves using a temperature sensor to collect real-time temperature data at a specific point within the meter box. This measurement reflects the overall thermal state within the meter box, encompassing the combined effects of various factors, including normal and abnormal heating, as well as ambient temperature.

[0092] Obtaining the measured current values of phases A, B, and C, as well as the neutral line, of the meter box's incoming power lines refers to collecting real-time current data on the phase conductors and neutral line of the meter box's power incoming power lines using current sensors. This current data is the fundamental cause of heat generation in the main heat sources inside the meter box (such as wire clamps and wires).

[0093] Calculating an expected box temperature based on the measured current values of phases A, B, and C, the neutral current, and a baseline thermal model representing the healthy state of the wire clamp means using current data as input and a pre-established mathematical model that simulates the thermal behavior of the meter box in a healthy state to predict the expected temperature inside the meter box under current load conditions.

[0094] The temperature deviation value between the calculated temperature measurement value and the expected box temperature value refers to the difference between the actually measured box temperature and the expected box temperature predicted by the reference thermal model.

[0095] Calculating the heating index values related to the current of the A-phase line, the B-phase line, the C-phase line, and the N-line line based on the measured current values of the A-phase line, the B-phase line, the C-phase line, and the neutral line respectively refers to calculating the index that quantifies the contribution of the current of each line to the total heating in the box based on the real-time current data of each phase and the neutral line.

[0096] Among them, calculating and tracking the dynamic correlation index between the time series of temperature deviation values and the time series of heating index values of phase A, phase B, phase C and N lines refers to analyzing the statistical correlation between the time series of temperature deviation representing abnormal temperature rise and the time series of heating index of each line load, and continuously monitoring the changes in this correlation.

[0097] Among them, according to the changing trend of the dynamic correlation index, when the dynamic correlation index corresponding to a specific line shows an enhancement and reaches the preset criterion conditions, it is determined that the wire clamp of the specific line has undergone early degradation. This means that based on the observation of the changes in the dynamic correlation indicators of each line over time, when the correlation index of a certain line continues to enhance and reaches the set threshold, it is considered that the contact resistance of the wire clamp of the line may be increasing, resulting in an increase in abnormal heat generation, and thus the wire clamp is determined to be in an early degradation state.

[0098] Among them, when the meter box operates in a low-load cycle, the temperature measurement values and the A-phase, B-phase, C-phase current measurement values, and the neutral line current measurement values obtained during the cycle are used to calibrate the parameters in the baseline thermal model. This means that during a time period when the meter box load is low and relatively stable, the temperature and current data collected at this time are used to adjust the parameters in the baseline thermal model so that its prediction results more accurately reflect the thermal behavior under a healthy state.

[0099] The core innovation of this application lies in the fact that, using only a single temperature sensor, a baseline thermal model is constructed to calculate the temperature deviation to highlight abnormal temperature rise, and the dynamic correlation between the temperature deviation and the heating index of each phase and neutral line current is further analyzed, thereby achieving the location and early warning of the wire clamp of which specific line inside the meter box has experienced early degradation. At the same time, a model calibration mechanism under low load cycles is introduced to improve the accuracy of monitoring.

[0100] The working principle of this technical solution is based on the following logic: the wire clamps inside the meter box generate heat due to the current they carry, and their heat generation power is proportional to the square of the current and the contact resistance. In the context of unbalanced three-phase load, the currents carried by the wire clamps of each phase and the neutral line differ. In a healthy state, the total temperature rise inside the box is primarily determined by the heat generated by the current flowing through each line through its nominal resistance and the ambient temperature. When the wire clamp of a phase (or neutral line) degrades prematurely, its contact resistance increases, causing additional heat to be generated in that line at the same current. Although the total temperature rise caused by this additional heat may be small and easily overwhelmed by interference, its generation pattern is specific: its magnitude varies directly with the square of the current flowing through that specific phase (or neutral line).

[0101] This solution measures the actual temperature using a temperature sensor placed inside the box. The expected temperature is calculated using real-time current data from each phase and neutral line, combined with a baseline thermal model that describes the relationship between heat generation and temperature under healthy conditions. The deviation (ΔT) between the actual and expected temperature reflects any factors not accurately described by the model, including model errors, environmental influences, and potential abnormal heating caused by degradation.

[0102] The core step is to continuously calculate the dynamic correlation coefficient between the time series of this temperature deviation ΔT and the time series of the squared current (I_phase²) for each line (A, B, C, N). Under normal circumstances, because ΔT primarily consists of random noise, model errors, or common pattern effects across all phases, its correlation with the squared current of a specific line is typically low and unstable. However, once the line clamp of a phase (for example, phase A) deteriorates, the additional heat generated (strongly correlated with I_A²) becomes a significant component of ΔT with a specific pattern, causing the correlation between ΔT and I_A² to significantly and continuously increase over time. By monitoring and comparing the changing trends of the correlation coefficients of the corresponding currents for these four lines (Corr(ΔT,I_A²), Corr(ΔT,I_B²), Corr(ΔT,I_C²), and Corr(ΔT,I_N²)), if one (and usually only one) increases significantly and exceeds a preset threshold, it can be determined that the line clamp on the corresponding line (phase or N) has experienced early degradation.

[0103] Through the above-mentioned solution, this application addresses the specific operating conditions where long-term three-phase load imbalance in the meter box causes differential electrothermal stresses on the phase and neutral clamps, leading to asynchronous early degradation. Furthermore, the application faces multiple challenges: severely limited monitoring resources, weak early fault signals, and the potential for interference from the complex thermal environment within the box. By utilizing only the existing phase current measurement data within the meter box and adding a single internal temperature sensor, this application effectively identifies and distinguishes which phase or neutral clamp is experiencing early degradation. By calculating the temperature deviation, the effects of ambient temperature and normal load heating on the measured temperature are filtered out, highlighting abnormal temperature rise signals. By analyzing the dynamic correlation between temperature deviation and heating indicators of each line and tracking their changing trends, abnormal heating synchronized with load changes on a specific line can be identified from weak abnormal signals, thereby locating the specific line experiencing early degradation. Using low-load cycles for model calibration improves the accuracy of the baseline thermal model, thereby enhancing the sensitivity and reliability of fault detection. This solution, requiring only the addition of a single temperature sensor, is low-cost and easily applicable.

[0104] In some of the above-mentioned schemes of the present application, a method is proposed to calculate and track the dynamic correlation index between the time series of the temperature deviation value between the temperature measurement value and the expected box temperature value and the time series of the heating index value related to the current of the A-phase line, the B-phase line, the C-phase line, and the N-line line to identify the early degradation of the wire clamp. However, in the actual operating environment, there may be instantaneous fluctuations in the internal temperature of the meter box and the line current, resulting in a short-term increase or fluctuation in the calculated dynamic correlation index. This instantaneous or short-term correlation enhancement may not be caused by the continuous degradation of the wire clamp, but by factors such as environmental interference or load transients. Simply making judgments based on the instantaneous value or short-term change trend of the dynamic correlation index is prone to false alarms or missed alarms, and cannot reliably distinguish between true signals caused by continuous degradation and interference signals caused by noise or transient events.

[0105] In this regard, the present application further proposes calculating a time series of dynamic correlation indicators between a time series of temperature deviation values and a time series of heating index values of each line; setting a discrimination threshold of the correlation indicator; setting a discrimination threshold of the duration; for the time series of the dynamic correlation indicator of a specific line, determining whether its value continuously exceeds the discrimination threshold of the correlation indicator within the time length defined by the discrimination threshold of the duration; when the time series of the dynamic correlation indicator of a specific line meets the continuous exceeding condition, identifying the enhancement trend presented by the dynamic correlation indicator as a real enhancement caused by continuous degradation.

[0106] Among them, this solution refines the steps of calculating and tracking the dynamic correlation index, aiming to identify the signal caused by the continuous deterioration of the wire clamp from the time series of the dynamic correlation index, and avoid misjudgment caused by instantaneous fluctuations or interference.

[0107] First, the time series of the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index value of each line is calculated. This provides a basis for the correlation data that changes over time for subsequent analysis. For example, the dynamic correlation index can be calculated using the sliding time window method. Within a sliding time window of a preset length, a time series segment of the temperature deviation value and a time series segment of the heating index value of a specific line are obtained within the window, and the correlation coefficient between the two segments, such as the Pearson correlation coefficient, is calculated. The correlation coefficient is used as the dynamic correlation index value corresponding to the position of the time window. By moving the time window according to the preset sliding step size and repeating the calculation, a time series of dynamic correlation indicators is formed.

[0108] Next, set the discrimination threshold of the correlation index and the discrimination threshold of the duration. The discrimination threshold of the correlation index is used to define the degree of correlation enhancement. For example, a value can be set. When the dynamic correlation index exceeds this value, it is considered that the correlation may be related to degradation. The discrimination threshold of the duration is used to define the length of time that such enhancement needs to last. For example, a time length can be set. When the dynamic correlation index continues to exceed the correlation index discrimination threshold for more than this length, the enhancement is considered to be continuous. These thresholds can be set based on historical data, expert experience, or through machine learning methods. For example, the discrimination threshold of the correlation index can be set to 0.7, and the discrimination threshold of the duration can be set to 24 hours.

[0109] Then, for the time series of the dynamic correlation index of a specific line, determine whether its value continuously exceeds the correlation index's discrimination threshold within the time length defined by the duration discrimination threshold. This step introduces the concepts of continuous exceeding and duration. By requiring that the dynamic correlation index not only exceed the correlation index's discrimination threshold, but also continuously remain above the threshold for a period of time exceeding the duration discrimination threshold, short-term, non-continuous correlation increases are effectively filtered out. For example, a counter or a timestamp record can be maintained. When the dynamic correlation index exceeds the correlation index's discrimination threshold, the timing or counting is started. If the index falls back below the threshold, the timing or counting is reset to zero. The continuous exceeding condition is only met when the continuous exceeding time reaches or exceeds the duration discrimination threshold. This persistence is a typical feature of early deterioration of the wire clamp, which causes abnormal heating and affects the temperature inside the box, because the degradation process is usually gradual and continuous.

[0110] Finally, when the time series of the dynamic correlation index of a specific line meets the continuous exceeding condition, the enhancing trend presented by the dynamic correlation index is identified as a true enhancement caused by continuous degradation. This identification step is based on the aforementioned continuity judgment, and the correlation enhancement signal that passes the duration test is confirmed to be associated with the real, continuous degradation of the wire clamp. By combining the enhancement of the dynamic correlation index with the duration requirement, this scheme can more accurately distinguish the real signal caused by the continuous degradation of the wire clamp from the instantaneous fluctuations caused by environmental interference or load transients, thereby improving the accuracy of fault monitoring. For example, if the dynamic correlation index of the phase A line exceeds 0.7 for 25 consecutive hours, and the duration discrimination threshold is set to 24 hours, it is considered that the phase A wire clamp has a real enhancement caused by continuous degradation. This identification provides a more reliable basis for subsequent fault judgment.

[0111] In some of the above-mentioned schemes of the present application, a method for monitoring the fault of the meter box wire clamp based on the dynamic correlation between temperature deviation and current heating index is proposed. This method relies on a baseline thermal model representing the health status of the wire clamp to calculate the expected temperature value inside the box. However, in actual operation, the ambient temperature of the meter box, the heat dissipation conditions of the box, and the heating conditions of other components in the box may change over time, resulting in a deviation between the baseline thermal model and the actual thermal environment. This model deviation will make the calculated expected box temperature value inaccurate, thereby affecting the accuracy of the temperature deviation value between the temperature measurement value and the expected box temperature value, and may ultimately interfere with the calculation and tracking of the dynamic correlation index between the temperature deviation and the heating index of each line, reduce the accuracy and reliability of fault monitoring, and may cause false alarms or missed alarms. Therefore, a method is needed to calibrate the parameters of the baseline thermal model in a timely and accurate manner to ensure that the model can reflect the current actual thermal environment and improve the accuracy of monitoring.

[0112] In this regard, the present application further proposes that when the meter box operates in a low-load cycle, the steps of calibrating the parameters in the reference thermal model using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values and the neutral current measurement values obtained during the cycle include:

[0113] Setting a first threshold value for defining a current amplitude of a low load;

[0114] setting a second threshold value for defining a current variation index of load stability and a length of a time window based on which the current variation index is calculated;

[0115] Obtain the current measurement values of phases A, B, and C of the incoming line of the meter box, as well as the neutral line current measurement value;

[0116] Determining whether the current measurement value is lower than a first threshold;

[0117] Calculate the current change index within the time window based on the current measurement value and the time window length;

[0118] determining whether the current change index is lower than a second threshold;

[0119] When the current measurement value is lower than a first threshold and the current change index is lower than a second threshold, determining the current operation cycle as a low-load cycle;

[0120] During a determined low-load cycle, parameters in the baseline thermal model are calibrated using temperature measurements and phase A, phase B, phase C current measurements, and neutral current measurements obtained during the cycle.

[0121] The method provides a basic criterion for identifying a suitable time for calibration by setting a first threshold for the current amplitude that defines low load. The current amplitude can be the maximum value of each phase current or the neutral current, or the average value of each phase current and the neutral current.

[0122] Furthermore, setting a second threshold for the current variation index used to define load stability and the length of the time window based on which the current variation index is calculated further refines the criteria for selecting the calibration timing. The current variation index can be the difference between the maximum and minimum current values within the time window, the standard deviation of the current, the rate of change of the current, etc.

[0123] The time window length can be set based on the actual application scenario and data sampling frequency, for example, from several minutes to several dozen minutes. The current measurements for phases A, B, and C of the incoming power lines to the meter box, as well as the neutral current, are obtained. These measurements serve as the foundational data for load determination and subsequent model calibration. Based on the obtained current measurements, a determination is made as to whether they are below a set first threshold, preliminarily identifying periods of low load.

[0124] Simultaneously, based on the current measurement and the set time window length, a current variation index within that time window is calculated, and a determination is made as to whether this index is below a set second threshold, thereby assessing the stability of the current load. When both the current measurement and the current variation index are below the first threshold and the second threshold, the current operating cycle is comprehensively determined to be a low-load, stable cycle suitable for model calibration. Within the determined low-load cycle, the parameters in the baseline thermal model are calibrated using the temperature measurements and the current measurements of phases A, B, and C, as well as the neutral current, acquired during that period.

[0125] The baseline thermal model can be a physical model based on the principles of a thermal network, consisting of heat source nodes, thermal resistance nodes, and thermal capacitance nodes. The calibration process can employ parameter estimation methods, such as least squares or optimization algorithms, using actual temperature and current data collected during a low-load cycle to adjust model parameters such as thermal resistance and capacitance to minimize the error between the expected temperature calculated by the model and the actual measured temperature.

[0126] By calibrating under low-load and stable conditions, the impact of the wire clamp's own heating (even if there is early degradation, its abnormal heating is relatively small) on the temperature inside the box can be reduced. This makes the influence of non-wire clamp heating factors, such as ambient temperature and box heat dissipation, relatively prominent and stable. At this time, using actual measurement data to calibrate the model can more accurately adjust the model parameters to better reflect the current actual thermal environment and the influence of non-wire clamp heating factors.

[0127] As a result, the expected in-box temperature value calculated by the calibrated benchmark thermal model will be more accurate, which directly improves the accuracy of the temperature deviation value between the temperature measurement value and the expected in-box temperature value. In the overall fault monitoring method, the accuracy of the temperature deviation value is the basis for calculating the dynamic correlation index between the temperature deviation value time series and the heating index value time series of each line. Accurate temperature deviation values and heating index value time series can make the calculated dynamic correlation index more realistically reflect the changes in the health status of the wire clamp. Therefore, by timely and accurately calibrating the benchmark thermal model, this scheme improves the reliability of temperature deviation and dynamic correlation calculations, thereby enhancing the accuracy and robustness of wire clamp fault monitoring and reducing the risk of false alarms or missed alarms.

[0128] In some of the above-mentioned schemes of the present application, it is proposed to calculate and track the dynamic correlation index between the time series of temperature deviation values and the time series of heating index values of each line to reflect the degradation state of the wire clamp. However, when calculating the correlation between time series, it is usually necessary to set a time window length. The load characteristics of the meter box are time-varying and random, and the load fluctuation characteristics of different lines are also different. At the same time, the degradation of the wire clamp is a gradual process. At different degradation stages, the correlation between its abnormal heating and current may be different. If a fixed time window length is used to calculate the dynamic correlation index, it may not be able to fully adapt to these changing working conditions and degradation stages. For example, when the load fluctuates violently, a window that is too long may smooth out important instantaneous correlation changes; when the amplitude of the early degradation signal is small, a window that is too short may cause unstable correlation calculation due to noise interference. Therefore, how to adaptively determine the time window length used to calculate the dynamic correlation index based on the actual operating status of the meter box, especially the historical fluctuation characteristics of the line load and the current degradation stage of the wire clamp, so as to improve the consistency of the correlation calculation with the actual situation and the ability to detect early degradation signals, is a technical problem that needs to be solved.

[0129] In this regard, the present application further proposes the steps of calculating and tracking the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the phase A line, the phase B line, the phase C line, and the N line, including:

[0130] Obtain the quantitative value of historical fluctuation characteristics of line load;

[0131] Obtain information about the identified degradation stage of the line;

[0132] Determine the length of the time window used to calculate the dynamic correlation index based on the quantified value of the historical fluctuation characteristics and the information of the identified degradation stages;

[0133] Using the determined time window length, a dynamic correlation index is calculated between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line.

[0134] Among them, this solution aims to solve the problem of how to adaptively select the appropriate time window length according to actual operating conditions when calculating the dynamic correlation between temperature deviation and line heating index.

[0135] First, a quantitative value of the historical load fluctuation characteristics is obtained. This provides information about the degree of current variation in the line over time. Load fluctuation characteristics affect the speed and magnitude of temperature response, thereby affecting the correlation between temperature deviation and current heating. This quantitative value of historical fluctuation characteristics can be calculated based on historical current measurement data. For example, the standard deviation, rate of change, or maximum change amplitude within a specific time period can be calculated to quantify the degree of load fluctuation.

[0136] Secondly, the current identified degradation stage of the line is obtained, which reflects the current state of the wire clamp degradation. At different degradation stages, the wire clamp's contact resistance may vary, and the relationship between its abnormal heating and current may also change. This degradation stage information can be identified based on an analysis of historical trends in dynamic correlation indicators. For example, different correlation thresholds or rate of change thresholds can be set to define different degradation stages, such as normal stage, early degradation stage, and mid-stage degradation stage.

[0137] Then, based on the obtained quantitative values of the historical fluctuation characteristics of the line load and the identified degradation stage information, the length of the time window used to calculate the dynamic correlation index is determined, and the historical operation data and current degradation status information are used to guide the selection of the correlation calculation window.

[0138] Furthermore, in order to adjust the window length more precisely, the change rate of the dynamic correlation index within a preset time period may be calculated, and an adjustment factor may be determined according to the change rate.

[0139] Finally, the obtained baseline time window length is calculated (e.g., multiplied or added) with the adjustment factor to obtain the time window length used to calculate the dynamic correlation index. This method eliminates the need for a fixed time window length and instead adjusts it based on actual conditions, making subsequent correlation calculations more targeted and effective.

[0140] Using the determined time window length, a dynamic correlation index is calculated between the time series of temperature deviation values and the time series of heating index values for each line. This calculation can be performed using a sliding time window method, with the sliding window length (i.e., the window length determined in the previous step) and the sliding step size set. By using a time window adaptively determined based on the current operating conditions and degradation stage, the calculated dynamic correlation index can more accurately reflect the true degree of correlation between temperature deviation and heating of a specific line current.

[0141] In some of the aforementioned solutions of this application, it is proposed to determine the time window length used to calculate the dynamic correlation index based on the quantified value of the historical fluctuation characteristics and the information of the identified degradation stage, in order to improve the accuracy of the correlation calculation. However, simply determining a fixed time window length based on these two pieces of information may not fully adapt to the dynamic changes in actual load fluctuations and degradation processes. As a result, under certain operating conditions or degradation stages, the determined time window length may be suboptimal, affecting the dynamic correlation index's ability to capture early degradation signals and its robustness to noise.

[0142] In this regard, the present application further proposes a method for determining the length of a time window for calculating a dynamic correlation index based on the quantified value of historical fluctuation characteristics and the identified degradation stage information. The method includes:

[0143] Definition of discrete levels of quantified values of historical fluctuation characteristics and discrete levels of information on identified degradation stages;

[0144] A two-dimensional lookup table is constructed, using the discrete levels of the historical fluctuation characteristic quantization value and the discrete levels of the identified degradation stage information as indexes, and a reference time window length value is preset in each cell of the two-dimensional lookup table;

[0145] Obtaining a reference time window length value from a two-dimensional lookup table according to the discrete level corresponding to the quantized value of the historical fluctuation characteristic and the discrete level corresponding to the identified degradation stage information;

[0146] Calculate the rate of change of the dynamic correlation index within a preset time period;

[0147] Determine an adjustment factor based on the rate of change of the dynamic correlation index;

[0148] The obtained reference time window length value is calculated with the adjustment factor to obtain the time window length used to calculate the dynamic correlation index.

[0149] The method first defines discrete levels for the quantified value of historical fluctuation characteristics and the identified degradation stage information. The quantified value of historical fluctuation characteristics can reflect the severity of changes in line load current over time. For example, it can be quantified based on the current standard deviation, coefficient of variation, or the energy of specific frequency components, and can be divided into three discrete levels: low, medium, and high. The identified degradation stage information indicates the current degradation state of the monitored object. For example, it can be divided into discrete levels such as healthy, early degradation, and medium degradation based on historical monitoring data or preset rules.

[0150] Through discretization, continuous or complex input information is converted into a limited, manageable set of categories. Next, a two-dimensional lookup table is constructed, with row indexes based on the discrete levels of historical fluctuation characteristic quantization values and column indexes based on the discrete levels of identified degradation stage information. Each cell in the table is assigned a baseline time window length. For example, a longer time window (e.g., 24 hours) might be used in low-volatility, healthy phases to enhance noise immunity; a shorter time window (e.g., 6 hours) might be used in high-volatility, early-stage degradation phases to increase sensitivity to rapid changes. These baseline values can be determined based on experience, simulation, or historical data analysis. The discrete levels of the current actual historical fluctuation characteristic quantization values and identified degradation stage information are then determined and used as indices to retrieve the corresponding baseline time window length values from the two-dimensional lookup table. This provides a preliminary, context-based time window setting.

[0151] Based on this, the method further calculates the rate of change of the dynamic correlation index within a preset time period. This rate of change can reflect the trend and speed of the correlation index's change over time. For example, the change in the correlation index's value within the last hour can be calculated and divided by the time period. Based on the calculated rate of change of the dynamic correlation index, an adjustment factor is determined. This adjustment factor is used to modify the length of the baseline time window. For example, when the rate of change of the dynamic correlation index is high, it may indicate signs of rapidly developing degradation. In this case, the adjustment factor can reduce the final time window length to capture signals more quickly. When the rate of change of the dynamic correlation index is low, it indicates a relatively stable state. In this case, the adjustment factor can increase the final time window length to improve calculation stability. The adjustment factor can be determined in various ways. For example, multiple ranges of rate of change values can be preset, and a corresponding adjustment factor value can be assigned to each range. The adjustment factor can be obtained by searching and determining the range to which it belongs. Alternatively, a function can be used to map the rate of change to the adjustment factor.

[0152] Finally, the obtained reference time window length value is calculated with the adjustment factor determined according to the correlation change rate to obtain the final time window length used to calculate the dynamic correlation index. The calculation method can be multiplication, addition or other combination methods. For example, the final window length can be equal to the reference window length multiplied by the adjustment factor, or equal to the reference window length plus a correction related to the adjustment factor. This method combines the reference value and dynamic adjustment, so that the determination of the time window length takes into account historical and stage factors, and can be dynamically optimized according to the real-time change trend of the correlation itself. This method is combined with the method of calculating the dynamic correlation index between the temperature deviation value and the heating index value time series. By using a time window that is more suitable for the current working conditions and degradation stage, the accuracy of the correlation calculation is improved, thereby enhancing the fault monitoring ability to capture early degradation signals and the robustness to noise.

[0153] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0154] First, the quantized historical fluctuation characteristics are discretized into multiple levels. For example, the standard deviation of the load current can be divided into three levels: low fluctuation, medium fluctuation, and high fluctuation. At the same time, the identified degradation stage information is also discretized into multiple levels. For example, the degradation stage can be divided into three levels: healthy, early warning, and warning.

[0155] Next, a two-dimensional lookup table is constructed, with row indices corresponding to discrete levels of historical fluctuation characteristic quantization values and column indices corresponding to discrete levels of identified degradation stage information. Each cell in the table is assigned a baseline time window length value. These values can be determined based on experience, simulation, or historical data analysis. For example, a longer time window (e.g., 24 hours) might be used during low-volatility / healthy phases, while a shorter time window (e.g., 6 hours) might be used during high-volatility / warning phases. Based on the current actual historical fluctuation characteristic quantization values and identified degradation stage information, their corresponding discrete levels are determined and used as indices to retrieve the corresponding baseline time window length values from the two-dimensional lookup table.

[0156] On this basis, the rate of change of the dynamic correlation index within a preset time period is calculated. For example, the current dynamic correlation index value and the value at the start of a preset time period (e.g., the past hour) can be obtained. The difference between the two can be calculated and divided by the length of the preset time period to obtain the rate of change. An adjustment factor is determined based on the calculated rate of change of the dynamic correlation index. For example, multiple ranges of rate of change values can be preset (e.g., rate of change less than 0, rate of change between 0 and 0.01, and rate of change greater than 0.01), with a corresponding adjustment factor value preset for each range (e.g., an adjustment factor of 1.2 for a rate of change less than 0, an adjustment factor of 1.0 for a rate of change between 0 and 0.01, and an adjustment factor of 0.8 for a rate of change greater than 0.01). The calculated rate of change is compared with these ranges to determine the range to which it belongs, and the corresponding preset adjustment factor value is selected.

[0157] Finally, the obtained reference time window length value is calculated and multiplied by the determined adjustment factor, for example, to obtain the final time window length used to calculate the dynamic correlation index. Therefore, the determination of the time window length comprehensively considers historical load characteristics, the current degradation state, and the dynamic change trend of the correlation index itself.

[0158] Through the above technical solution, the present application solves the problem that simply determining a fixed time window length based on the quantitative value of historical fluctuation characteristics and the identified degradation stage information cannot fully adapt to the actual load fluctuations and dynamic changes in the degradation process. By introducing the change rate of the dynamic correlation index itself as the basis for adjustment, the determined time window length can be dynamically adjusted according to the real-time change trend of the correlation index. When the correlation index changes rapidly, the time window can be shortened, thereby improving the sensitivity of capturing early degradation signals; when the correlation index changes slowly, the time window can be lengthened, thereby enhancing the robustness to noise. This dynamic adaptability improves the accuracy of the calculation of the dynamic correlation index, thereby improving the accuracy and sensitivity of fault monitoring.

[0159] The present application proposes the steps of determining an adjustment factor based on the change rate of a dynamic correlation index, including: pre-setting multiple numerical intervals of the change rates of the dynamic correlation index; pre-setting a corresponding adjustment factor value for each numerical interval; comparing the calculated change rate of the dynamic correlation index with multiple numerical intervals to determine the numerical interval to which the change rate of the dynamic correlation index belongs; and selecting the preset adjustment factor value corresponding to the numerical interval to which the change rate of the dynamic correlation index belongs as the determined adjustment factor.

[0160] This solution addresses the problem of determining an adjustment factor based on the rate of change of a dynamic correlation index by proposing a technique based on numerical interval search. First, by presetting multiple numerical intervals for the rate of change of the dynamic correlation index, the continuously changing rate of change of the dynamic correlation index is discretized into different levels or states. For example, the rate of change can be set to multiple intervals, such as less than a negative threshold (indicating a rapid decline), between a negative threshold and zero (indicating a slow decline), close to zero (indicating stability), between zero and a positive threshold (indicating a slow increase), and greater than a positive threshold (indicating a rapid increase).

[0161] Next, a corresponding adjustment factor value is pre-set for each set numerical interval, which establishes a clear, segmented mapping relationship, that is, the change rate in different ranges corresponds to different adjustment factors. For example, for a rapidly declining interval, an adjustment factor less than 1 can be set; for an interval close to zero, an adjustment factor equal to 1 can be set; for a rapidly rising interval, an adjustment factor greater than 1 can be set. Then, the actual calculated rate of change of the dynamic correlation index is compared with these preset numerical intervals to determine which interval the current rate of change falls within. Finally, the preset adjustment factor value corresponding to the numerical interval to which the rate of change belongs is selected as the final adjustment factor.

[0162] Furthermore, the determined adjustment factor is used to adjust the length of a base time window, thereby obtaining the time window length used to calculate the dynamic correlation index. The base time window length can be determined based on the historical fluctuation characteristics of the line load and the identified degradation stage information. The time window length can be dynamically adjusted by calculating (e.g., multiplying) the base time window length with the adjustment factor determined based on the rate of change of the dynamic correlation index. When the rate of change of the dynamic correlation index is high, it indicates that the correlation between the temperature deviation and the heat generation index is rapidly changing. In this case, the time window length may need to be shortened to more quickly capture such changes and improve the real-time monitoring performance. When the rate of change is low, it indicates that the correlation is relatively stable. In this case, the time window length can be appropriately extended to smooth the data, reduce noise interference, and improve the stability of the correlation calculation. This method of adjusting the calculation time window length based on the rate of change of the correlation index makes the tracking process of the dynamic correlation index more adaptable, better balancing real-time performance and interference resistance, and thus more accurately reflects the degradation status of the wire clamp.

[0163] Specifically, the process of determining the adjustment factor first requires pre-defining a series of non-overlapping numerical intervals to cover the possible range of change rates of the dynamic correlation index. For example, the intervals can be set as (-∞, -0.02], (-0.02, -0.005], (-0.005, 0.005), [0.005, 0.02), [0.02, +∞). Subsequently, a corresponding adjustment factor value is assigned to each interval. These values can be determined based on experience or experiments, for example, corresponding to 0.8, 0.9, 1.0, 1.1, and 1.2 respectively. When the system is running, the rate of change of the current dynamic correlation index within a preset time period is calculated. The calculated rate of change is compared with the preset numerical interval to determine which interval it falls into. For example, if the calculated rate of change is 0.015, it falls into the interval [0.005, 0.02). Therefore, the preset adjustment factor value corresponding to this interval, i.e. 1.1, is selected as the adjustment factor determined this time.

[0164] In this way, this solution provides a specific and operational method to determine the adjustment factor based on the rate of change of the dynamic correlation index, making the adjustment process of the time window length of the dynamic correlation index more standardized and controllable. It can systematically select the appropriate adjustment factor according to the speed of change of the dynamic correlation index, thereby more accurately determining the time window length used to calculate the dynamic correlation index, thereby improving the adaptability and accuracy of fault monitoring.

[0165] The present application further proposes that the steps of calculating the rate of change of the dynamic correlation index within a preset time period include:

[0166] Obtain the dynamic correlation index value corresponding to the start time of the preset time period;

[0167] Get the dynamic correlation index value corresponding to the end time of the preset time period;

[0168] Calculate the numerical difference between the dynamic correlation index value at the end time and the dynamic correlation index value at the start time;

[0169] The numerical difference is divided by the length of the preset time period to obtain the rate of change of the dynamic correlation index within the preset time period.

[0170] Specifically, to calculate the rate of change of the dynamic correlation index within a preset time period, we first need to determine a preset time period. For example, the preset time period can be set to the past 24 hours. Then, we obtain the dynamic correlation index value corresponding to the start time of the preset time period, that is, 24 hours ago. This value is read from the historical records.

[0171] Next, the dynamic correlation index value corresponding to the end time of the preset time period, that is, the current time, is obtained. This value is the most recently calculated value.

[0172] Then, the difference between the current dynamic correlation index value and the dynamic correlation index value 24 hours ago is calculated. For example, if the current value is 0.8 and the value 24 hours ago was 0.6, the difference is 0.2.

[0173] Finally, divide this difference by the length of the preset time period, i.e., 24 hours. This yields the average rate of change of the dynamic correlation index over the past 24 hours: for example, 0.2 / 24 ≈ 0.0083 per hour. This calculation provides a standardized method for quantifying the speed and direction of change in the dynamic correlation index.

[0174] By adopting this clear calculation procedure, errors introduced by inconsistent or inaccurate calculation methods are avoided. The accurately calculated rate of change is then used to determine an adjustment factor, which is used to adjust the length of the time window for calculating the dynamic correlation index. For example, a lookup table can be preset to map different rate of change value intervals to different adjustment factors. A higher positive rate of change may correspond to an adjustment factor that shortens the time window to capture degradation trends more quickly; a rate of change close to zero may correspond to an adjustment factor that maintains or slightly extends the time window to improve calculation stability.

[0175] In this way, the rate of change of the dynamic correlation index is accurately quantified and used as a reliable basis to dynamically adjust the time window, so that the calculation of the dynamic correlation index can better adapt to the actual degradation development speed, thereby improving the accuracy and responsiveness of monitoring.

[0176] In some of the above-mentioned schemes of the present application, it is proposed to calculate an expected box temperature value based on the A-phase, B-phase, and C-phase current measurements of the meter box incoming line, the neutral line current measurement value, and a baseline thermal model representing the health status of the wire clamp, and use the deviation between the expected box temperature value and the actual temperature measurement value to monitor the wire clamp fault. However, the internal structure of the meter box is complex, containing a variety of heat-generating components, and its heat exchange with the external environment is affected by a variety of factors, resulting in a complex and changeable thermal environment in the box. A simple baseline thermal model may be difficult to accurately capture the temperature variation pattern in the box under different load conditions and environmental changes, especially it is difficult to accurately predict the temperature in a healthy state, so that the calculated temperature deviation value is not accurate enough, affecting the reliability of the subsequent fault judgment link, and may lead to false alarms or omissions of early wire clamp degradation. Therefore, how to construct a baseline thermal model that can more accurately reflect the thermal behavior of the meter box in a healthy state to improve the accuracy of the expected box temperature calculation is a technical problem that needs to be solved.

[0177] In this regard, the present application further proposes steps for calculating an expected temperature value inside the box based on the current measurement values of phases A, B, and C, the neutral line current measurement value, and a baseline thermal model representing the health status of the wire clamp, including: using a physical model based on the thermal circuit network principle as a baseline thermal model; calculating the healthy heating power of each line based on the current measurement values of phases A, B, and C, the neutral line current measurement value, and the preset contact resistance values of each line in the health status, and setting the healthy heating power as the heat input of the heat source node in the baseline thermal model; setting a thermal resistance node in the baseline thermal model that represents the heat transfer path inside the meter box and the heat dissipation path between the meter box and the external environment; setting a node in the baseline thermal model that represents the heat capacity of the meter box and its internal components; solving the baseline thermal model including the heat source node, the thermal resistance node, and the heat capacity node to obtain the expected temperature value inside the box.

[0178] A physical model based on the principle of a heat circuit network is used as the baseline thermal model. This model, based on the laws of physics, simulates the generation, transfer, and storage of heat within the meter box. For example, the model can contain multiple nodes, each representing the temperature of a specific area or component within the meter box. Nodes are connected by thermal resistances, representing the heat transfer paths. The magnitude of the thermal resistance depends on the thermal conductivity of the material, the geometry, and the heat transfer method (conduction, convection, or radiation).

[0179] Furthermore, the healthy heating power of each line under the current load is calculated based on the actual measured current values of phases A, B, and C and the neutral line current value, as well as the preset contact resistance value of each line in a healthy state. The heating power can be calculated according to Joule's law. For example, for a phase line, its healthy heating power is equal to the square of the measured current value of the phase multiplied by the preset healthy contact resistance value. These calculated healthy heating powers are set as the heat input of the heat source node in the baseline thermal model. For example, the model can have one or more heat source nodes, corresponding to the heat generation of each phase and neutral line clamp in a healthy state.

[0180] In the baseline thermal model, thermal resistance nodes are defined to represent the heat transfer paths within the meter box and the heat dissipation paths between the meter box and the external environment. These resistance nodes quantify the resistance to heat transfer from the heat source to the air inside the box, the box body, and heat dissipation from the box body to the external environment. For example, thermal resistances can be defined to represent convection from the wire clamp to the box air, convection from the box air to the box body, conduction from the box material, and convection and radiation from the box surface to the external environment. Thermal resistance values can be obtained through theoretical calculation or experimental calibration.

[0181] In the baseline thermal model, nodes representing the heat capacity of the meter box and its internal components are set. Heat capacity reflects an object's ability to store heat and influences the dynamic response of temperature changes over time. For example, heat capacity nodes can be set to represent the heat capacity of major heat-generating or heat-storing components within the box, such as the air, meter, circuit breaker, and wire clamps. Heat capacity values can be calculated using material density, specific heat, and volume, or obtained through experimental calibration.

[0182] Finally, the baseline thermal model, which includes heat source nodes, thermal resistance nodes, and thermal capacitance nodes, is solved to obtain the expected temperature inside the box. This solution can be a steady-state analysis (ignoring the effects of thermal capacitance and calculating the steady-state temperature) or a transient analysis (accounting for the effects of thermal capacitance and calculating the temperature change over time). By solving this physical model, we can obtain a predicted temperature at a specific location inside the meter box (for example, where the temperature sensor is located) when the meter box is in a healthy state under the current load.

[0183] Therefore, by adopting a thermal circuit network model based on physical principles and combining it with actual current input and health status parameters, it is possible to more accurately simulate the thermal behavior of the meter box and predict the temperature inside the box in a healthy state. When this more accurate expected temperature value is compared with the actual measured temperature value inside the box, the resulting temperature deviation value can more effectively reflect whether there is actually an abnormal temperature rise, thereby improving the accuracy of the subsequent fault determination link. For example, when the actual temperature measurement value is significantly higher than the accurately predicted healthy state temperature value, the larger deviation value is more likely to indicate abnormal heating rather than normal temperature changes caused by inaccurate model predictions or environmental fluctuations. This accurate deviation value calculation provides a reliable foundation for subsequently locating the fault line by analyzing the correlation between the deviation value and the current heating indicators of each line.

[0184] In some of the above-mentioned schemes of the present application, it is proposed to calculate and track the dynamic correlation index between the time series of temperature deviation values and the time series of heating index values of each line to determine the wire clamp fault according to their changing trends. However, in actual applications, how to specifically and effectively calculate and track this dynamic correlation index so that it can accurately reflect the enhanced dynamic correlation between the weak abnormal heating caused by the early deterioration of the wire clamp and the current load, and overcome the influence of ambient temperature fluctuations, interference from other heat sources in the box, and signal noise, is a key challenge to achieve reliable fault monitoring.

[0185] In this regard, the present application further proposes the steps of calculating and tracking the dynamic correlation index between the time series of temperature deviation values and the time series of heating index values of the A-phase line, the B-phase line, the C-phase line and the N-line line, including: setting the length of the sliding time window and the sliding step for calculating the dynamic correlation index; at each sliding time window position, obtaining a time series segment of the temperature deviation value within the sliding time window; obtaining a time series segment of the heating index value of any one of the A-phase line, the B-phase line, the C-phase line or the N-line line corresponding to the sliding time window position; calculating the Pearson correlation coefficient between the obtained time series segment of the temperature deviation value and the obtained time series segment of the heating index value of the specific line; assigning the calculated Pearson correlation coefficient to the dynamic correlation index value of the specific line corresponding to the sliding time window position; as the sliding time window moves according to the sliding step, repeatedly executing the steps of obtaining the time series segment of the temperature deviation value, obtaining the time series segment of the heating index value of the specific line, calculating the Pearson correlation coefficient and assigning the dynamic correlation index value to form the dynamic correlation index of each specific line.

[0186] This solution provides a specific method for calculating and tracking the dynamic correlation between temperature deviations and heating index values for each line. This approach aims to address the problem of effectively quantifying and tracking this dynamic relationship to support early fault diagnosis. By employing a method based on a sliding time window and the Pearson correlation coefficient, this solution can capture the increasing correlation between abnormal temperature rise caused by early deterioration of the wire clamp and the current load, even if the abnormal temperature rise signal is weak and affected by environmental interference.

[0187] Specifically, first, the length of the sliding time window and the sliding step used to calculate the dynamic correlation index define the time scale and update frequency for analyzing dynamic correlation. The length of the sliding time window determines the amount of historical data considered each time the correlation is calculated, affecting the sensitivity to short-term fluctuations and the degree of smoothness of long-term trends; the sliding step determines the frequency of updating the correlation index, affecting the precision of dynamic tracking. For example, the length of the sliding time window can be set to a data volume of several hours to several days, and the sliding step can be set to several minutes to several hours. By reasonably setting these two parameters, the calculated correlation index can better reflect the gradual process caused by the deterioration of the wire clamp.

[0188] Next, at each sliding time window position, a time series segment of the temperature deviation values within that sliding time window is obtained, along with a time series segment of the heating index values for any of the phases A, B, C, or N corresponding to that sliding time window position. This ensures that both temperature deviation data and heating index data for a specific line are extracted within the same time window. This serves as the foundation for correlation analysis and ensures the temporal correspondence of the analysis objects. The time series of temperature deviation values is calculated from the measured temperature values and the expected in-chamber temperature values, while the time series of heating index values is calculated from the measured current values for each phase or neutral line.

[0189] Next, calculating the Pearson correlation coefficient between the acquired time series of temperature deviation values and the acquired time series of heating index values for a specific line is a key step in quantifying the correlation. The Pearson correlation coefficient is a statistic used to measure the strength and direction of the linear correlation between two sets of data. Here, calculating the Pearson correlation coefficient between the temperature deviation (representing abnormal temperature rise) and the line heating index (representing normal and potentially abnormal heating caused by current load) within a specific time window effectively quantifies the extent to which the temperature deviation within the box changes with changes in line current. For healthy clamps, temperature deviation is primarily affected by environmental and other factors and has a weak correlation with the current of a single line. However, when a line clamp deteriorates, resulting in increased contact resistance, the additional heat generated significantly increases with current, strengthening the correlation between the temperature deviation and the heating index related to that line's current. The Pearson correlation coefficient ranges from -1 to 1, with values close to 1 indicating a strong positive correlation, -1 indicating a strong negative correlation, and 0 indicating a weak correlation. The degradation of the wire clamp is usually manifested as an increase in contact resistance, which causes the heat generation power to increase with the square of the current, which will make the temperature deviation and the current-related heat generation index show a positive correlation and increase.

[0190] The calculated Pearson correlation coefficient is assigned as the dynamic correlation index value of the specific line corresponding to the position of the sliding time window, which clarifies the specific numerical source of the dynamic correlation index, that is, the Pearson correlation coefficient calculated in each time window.

[0191] Finally, as the sliding time window moves according to the sliding step size, the steps of obtaining a time series segment of temperature deviation values, obtaining a time series segment of heating index values for a specific line, calculating the Pearson correlation coefficient, and assigning the value to a dynamic correlation index are repeated to form a dynamic correlation index for each specific line, describing the dynamic tracking process. By continuously moving the sliding window and repeating the calculation, the system generates a time-varying sequence of correlation indicators for each line. Tracking the changing trends of this sequence, particularly its strengthening trends, can effectively indicate whether the wire clamps of a specific line are experiencing early degradation, thereby enabling early, localizable fault detection.

[0192] Secondly, refer to Figure 2 , the present application further proposes a meter box wire clamp fault monitoring system, the system comprising:

[0193] Temperature acquisition module 1, used to obtain the temperature measurement value of a single location in the meter box;

[0194] Current acquisition module 2, used to obtain the A-phase, B-phase, C-phase current measurement values of the meter box incoming line and the neutral line current measurement value;

[0195] An expected temperature calculation module 3 is configured to calculate an expected internal temperature value of the box based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and a reference thermal model representing the health status of the wire clamp;

[0196] The temperature deviation calculation module 4 is used to calculate the temperature deviation between the temperature measurement value and the expected temperature value in the box;

[0197] A heating index calculation module 5 is used to calculate the current-related heating index values of the A-phase line, the B-phase line, the C-phase line and the N-line line based on the A-phase, B-phase and C-phase current measurement values and the neutral line current measurement value;

[0198] A correlation calculation and tracking module 6 is used to calculate and track the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A phase line, the B phase line, the C phase line and the N line;

[0199] The fault determination module 7 is configured to determine that the wire clamp of a specific line has experienced early degradation when the dynamic correlation index corresponding to a specific line shows an enhancement and reaches a preset judgment condition based on the change trend of the dynamic correlation index;

[0200] The model calibration module 8 is used to calibrate the parameters in the reference thermal model when the meter box operates in a low load cycle using the temperature measurement values and the A-phase, B-phase, C-phase current measurement values and the neutral current measurement values obtained in the cycle.

[0201] By utilizing a single temperature sensor and existing current measurement data, combined with a baseline thermal model and dynamic correlation analysis, it is possible to effectively identify and locate early degradation faults of specific phase or neutral line clamps in the meter box. This approach has the advantages of low cost, the ability to effectively identify and distinguish which phase or neutral line clamp has shown signs of early degradation, strong anti-interference ability, sensitivity to early fault signals, and an adaptive model.

[0202] In addition, in some preferred embodiments, the meter box wire clamp fault monitoring system proposed in this application can perform any one of the steps in the above method.

[0203] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for monitoring faults of electric meter box wire clamps, characterized in that: include: Obtaining a temperature measurement value of a single location in the electric meter box; Obtaining the current measurement values of phase A, phase B, phase C and neutral line of the meter box incoming line; Calculating an expected temperature value inside the box based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and a baseline thermal model representing the health status of the clamp; Calculating a temperature deviation between the temperature measurement value and the expected temperature value in the box; Based on the measured current values of the phases A, B, and C and the measured current value of the neutral line, respectively, calculating the current-related heating index values of the phase A line, the phase B line, the phase C line, and the neutral line; Calculate and track a dynamic correlation index between a time series of the temperature deviation value and a time series of heating index values of each of the A-phase line, the B-phase line, the C-phase line, and the N-line line; According to the changing trend of the dynamic correlation index, when the dynamic correlation index corresponding to a specific line shows an enhancement and reaches a preset judgment condition, it is determined that the wire clamp of the specific line has undergone early degradation; When the electric meter box operates in a low-load cycle, calibrating parameters in the reference thermal model using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values, and the neutral current measurement value obtained during the cycle; The step of calculating an expected temperature value inside the box based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and a baseline thermal model representing the health status of the clamp includes: A physical model based on the heat circuit network principle is used as the reference thermal model; Calculating the healthy heating power of each line based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and the preset contact resistance value of each line in a healthy state, and setting the healthy heating power as the heat input of the heat source node in the benchmark thermal model; Setting thermal resistance nodes representing the heat transfer path inside the meter box and the heat dissipation path between the meter box and the external environment in the benchmark thermal model; Setting nodes representing the thermal capacity of the electric meter box and its internal components in the baseline thermal model; The benchmark thermal model including the heat source node, the thermal resistance node, and the heat capacity node is solved to obtain the expected temperature value in the box.

2. A method for monitoring faults of electric meter box wire clamps according to claim 1, characterized in that: The step of calculating and tracking the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line comprises: Calculate the time series of the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index value of each line; Set the discrimination threshold of the correlation index; Set the duration threshold; For the time series of the dynamic correlation index of a specific line, determining whether its value continuously exceeds the discrimination threshold of the correlation index within a time length defined by the discrimination threshold of the duration; When the time series of the dynamic correlation index of the specific line meets the continuous exceeding condition, the enhancement trend presented by the dynamic correlation index is recognized as a real enhancement caused by continuous degradation.

3. The method for monitoring a fault of an electric meter box wire clamp according to claim 1, characterized in that: The step of calibrating the parameters in the reference thermal model using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values and the neutral current measurement value obtained during the low-load cycle of the electric meter box includes: Setting a first threshold value for defining a current amplitude of a low load; setting a second threshold for defining a current variation indicator of load stability and a length of a time window based on which the current variation indicator is calculated; Obtaining the current measurement values of phase A, phase B, phase C and neutral line of the meter box incoming line; Determining whether any of the A-phase, B-phase, and C-phase current measurement values and the neutral line current measurement value are all lower than the first threshold; Based on the current measurement value and the time window length, respectively calculate the current change index of the A-phase line, the B-phase line, the C-phase line, and the N-line line within the time window length; Determine whether the current change index of each of the A-phase line, the B-phase line, the C-phase line, and the N-line line is lower than the second threshold; When the current measurement values are all lower than the first threshold and the current change indicators are all lower than the second threshold, determining the current operation cycle as the low-load cycle; During the determined low-load period, the parameters in the reference thermal model are calibrated using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values and the neutral current measurement value obtained during the period.

4. The method for monitoring a fault of an electric meter box wire clamp according to claim 1, characterized in that: The step of calculating a dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line comprises: Obtain the quantitative value of historical fluctuation characteristics of line load; Obtain information about the identified degradation stage of the line; Determining a time window length for calculating the dynamic correlation index based on the historical fluctuation characteristic quantified value and the identified degradation stage information; The determined time window length is used to calculate a dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line.

5. The method for monitoring a fault of an electric meter box wire clamp according to claim 4, characterized in that: The step of determining the length of the time window for calculating the dynamic correlation index based on the historical fluctuation characteristic quantization value and the identified degradation stage information includes: defining discrete levels of the quantized value of the historical fluctuation characteristic and discrete levels of the identified degradation stage information; Constructing a two-dimensional lookup table, using the discrete levels of the historical fluctuation characteristic quantization value and the discrete levels of the identified degradation stage information as indexes, and presetting a reference time window length value in each cell of the two-dimensional lookup table; Obtaining the reference time window length value from the two-dimensional lookup table according to the discrete level corresponding to the quantized value of the historical fluctuation characteristic and the discrete level corresponding to the identified degradation stage information; Calculating the rate of change of the dynamic correlation index within a preset time period; determining an adjustment factor according to the rate of change of the dynamic correlation index; Calculating the obtained reference time window length value and the adjustment factor to obtain a time window length for calculating the dynamic correlation index; The step of determining an adjustment factor according to the change rate of the dynamic correlation index includes: Presetting a plurality of numerical ranges of the dynamic correlation index change rates; Presetting a corresponding adjustment factor value for each of the numerical intervals; Comparing the calculated change rate of the dynamic correlation index with the multiple numerical intervals to determine the numerical interval to which the change rate of the dynamic correlation index belongs; A preset adjustment factor value corresponding to the numerical range to which the rate of change of the dynamic correlation index belongs is selected as the determined adjustment factor.

6. A method for monitoring faults of electric meter box wire clamps according to claim 5, characterized in that: The step of calculating the change rate of the dynamic correlation index within a preset time period includes: Obtaining the dynamic correlation index value corresponding to the start time of the preset time period; Obtaining the dynamic correlation index value corresponding to the end time of the preset time period; Calculating the numerical difference between the dynamic correlation index value at the end time and the dynamic correlation index value at the start time; The numerical difference is divided by the length of the preset time period to obtain the change rate of the dynamic correlation index within the preset time period.

7. The method for monitoring faults of electric meter box wire clamps according to claim 1, characterized in that: The step of calculating and tracking the dynamic correlation index between the time series of the temperature deviation value and the time series of the heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line comprises: Setting the length of the sliding time window and the sliding step size for calculating the dynamic correlation index; At each sliding time window position, obtaining a time series segment of the temperature deviation value within the sliding time window; Obtaining a time series segment of a heating index value of any one of the A-phase line, the B-phase line, the C-phase line, or the N-line line corresponding to the position of the sliding time window; Calculating a Pearson correlation coefficient between the acquired time series segment of the temperature deviation value and the acquired time series segment of the heating index value of the specific circuit; Assigning the calculated Pearson correlation coefficient as a dynamic correlation index value of a specific line corresponding to the position of the sliding time window; As the sliding time window moves according to the sliding step, the steps of obtaining the time series segment of the temperature deviation value, obtaining the time series segment of the heating index value of the specific line, calculating the Pearson correlation coefficient, and assigning the value to the dynamic correlation index are repeated to form the dynamic correlation index of each specific line.

8. A meter box wire clamp fault monitoring system, characterized in that: The system includes: A temperature acquisition module, used to obtain a temperature measurement value of a single location in the meter box; A current acquisition module is used to obtain the current measurement values of phase A, phase B, phase C and neutral line of the meter box incoming line; an expected temperature calculation module, configured to calculate an expected internal temperature value of the box based on the measured current values of the phases A, B, and C, the measured current value of the neutral line, and a reference thermal model representing the health status of the clamp; a temperature deviation calculation module, configured to calculate a temperature deviation between the temperature measurement value and the expected temperature value inside the box; a heating index calculation module, configured to calculate current-related heating index values of the A-phase line, the B-phase line, the C-phase line, and the N-line line based on the A-phase, B-phase, and C-phase current measurement values and the neutral line current measurement value; a correlation calculation and tracking module, configured to calculate and track a dynamic correlation index between a time series of the temperature deviation value and a time series of heating index values of each of the A-phase line, the B-phase line, the C-phase line, and the N-line line; a fault determination module, configured to determine, based on a change trend of the dynamic correlation index, that the wire clamp of a specific line has experienced early degradation when the dynamic correlation index corresponding to the specific line shows an enhancement and reaches a preset judgment condition; a model calibration module, configured to calibrate parameters in the reference thermal model using the temperature measurement values and the A-phase, B-phase, and C-phase current measurement values, and the neutral current measurement value obtained during a low-load cycle when the electricity meter box operates in the cycle; The step of calculating an expected temperature value inside the box based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and a baseline thermal model representing the health status of the clamp includes: A physical model based on the heat circuit network principle is used as the reference thermal model; Calculating the healthy heating power of each line based on the measured current values of phases A, B, and C, the measured current value of the neutral line, and the preset contact resistance value of each line in a healthy state, and setting the healthy heating power as the heat input of the heat source node in the benchmark thermal model; Setting thermal resistance nodes representing the heat transfer path inside the meter box and the heat dissipation path between the meter box and the external environment in the benchmark thermal model; Setting nodes representing the thermal capacity of the electric meter box and its internal components in the baseline thermal model; The benchmark thermal model including the heat source node, the thermal resistance node, and the heat capacity node is solved to obtain the expected temperature value in the box.

Images