Power metering box fault diagnosis method and system based on internet of things

Abstract

The application discloses a power metering box fault diagnosis method and system based on the Internet of Things, relates to the technical field of power equipment state monitoring and fault diagnosis, and comprises the following steps: identifying a load disturbance event according to current change characteristics in the operation process of the power metering box and determining a load disturbance occurrence time; determining a temperature rise response time caused by the load disturbance based on the load disturbance occurrence time; constructing a dynamic delay characteristic between the current disturbance and the temperature rise response according to a time difference between the load disturbance occurrence time and the temperature rise response time; matching the dynamic delay characteristic with a corresponding thermal response characteristic in a normal operation state of the power metering box, and judging whether the power metering box has an abnormal heating fault according to a matching result; and the application solves the problem that the existing power metering box fault diagnosis method does not consider the thermal inertia delay relationship between current change and temperature response, leading to the problems of easy misjudgment or missed detection of the abnormal heating fault, low diagnosis accuracy and reliability.

Description

Technical Field

[0001] This invention relates to the field of power equipment condition monitoring and fault diagnosis technology, and more specifically, to a method and system for fault diagnosis of power metering boxes based on the Internet of Things. Background Technology

[0002] With the development of IoT technology, electricity metering boxes are gradually becoming more intelligent. Existing IoT-based electricity metering boxes are typically equipped with current acquisition modules, voltage acquisition modules, temperature detection modules, and communication modules. By collecting electrical and environmental parameters during the operation of the metering box and uploading the collected data to a monitoring platform, the operating status of the metering box can be monitored and faults can be diagnosed.

[0003] Existing fault diagnosis methods for electricity metering boxes typically determine the operating status of the equipment based on changes in parameters such as current, voltage, power, and temperature. For example, they determine whether the metering box is faulty by judging abnormal changes in current or whether the temperature exceeds a preset threshold.

[0004] However, existing fault diagnosis methods mainly rely on parameter values ​​collected at the current moment for judgment, usually assuming a synchronous response relationship between changes in electrical parameters and temperature, without considering the thermal inertia effect that exists after load changes inside the metering box. In actual operation, when the load changes, the current change will occur before the temperature change inside the box, and the temperature rise at the terminals, conductive parts, etc., requires a certain period of heat accumulation before it becomes apparent, resulting in a time delay between the current change and the temperature response.

[0005] Therefore, when the metering box has early faults such as increased contact resistance or loose wiring, the resulting temperature rise may lag behind the current change. Existing methods cannot accurately distinguish between the transient temperature rise caused by normal load fluctuations and the continuous heating process caused by the fault, which can easily lead to misjudgment or missed detection of faults, reducing the accuracy and reliability of fault diagnosis of the power metering box. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a fault diagnosis method and system for power metering boxes based on the Internet of Things. By constructing an electrothermal coupling analysis method based on the dynamic delay characteristics between the moment of load disturbance and the moment of temperature rise response, the present invention solves the problem that the existing fault diagnosis methods for power metering boxes do not consider the thermal inertia delay relationship between current change and temperature response, which leads to the easy misjudgment or missed detection of abnormal heating faults and low diagnostic accuracy and reliability.

[0007] To achieve the above objectives, the present invention provides the following technical solution: Firstly, this application provides a fault diagnosis method for an electricity metering box based on the Internet of Things. The method includes: identifying load disturbance events based on the current change characteristics during the operation of the electricity metering box, and determining the time of occurrence of the load disturbance; analyzing the temperature response process within a preset time range based on the time of occurrence of the load disturbance, and determining the temperature rise response time caused by the load disturbance; constructing a dynamic delay feature between the current disturbance and the temperature rise response based on the time difference between the time of occurrence of the load disturbance and the time of temperature rise response; matching the dynamic delay feature with the corresponding thermal response feature under normal operation of the electricity metering box, and determining whether the electricity metering box has an abnormal heating fault based on the matching result.

[0008] In one embodiment, identifying load disturbance events based on current change characteristics during the operation of the power metering box and determining the time of occurrence of the load disturbance includes: collecting continuous current monitoring data, constructing a current change time series dataset according to the sampling time sequence, and extracting current disturbance change characteristics; matching the current disturbance change characteristics with preset disturbance judgment conditions, identifying a load disturbance event when the preset disturbance judgment conditions are met, and using the starting sampling time of the continuous current change interval that meets the conditions as the candidate occurrence time of the load disturbance; performing time correction on the candidate occurrence time of the load disturbance to determine the time of occurrence of the load disturbance.

[0009] In one embodiment, time correction is performed on the candidate occurrence time of load disturbance to determine the occurrence time of load disturbance. This includes: when the candidate occurrence time of load disturbance does not meet the preset disturbance judgment conditions, searching sequentially along the current sampling time sequence, and determining the sampling time that first simultaneously meets all the judgment conditions of current change amplitude, change rate and duration as the corrected disturbance occurrence time; using the corrected disturbance occurrence time as a reference, performing consistency verification on the current change data within the subsequent preset time window; if the verification passes, outputting the time as the final disturbance occurrence time; if the verification fails, continuing to search sequentially for the next sampling time that meets all the judgment conditions until the consistency verification is passed.

[0010] In one embodiment, determining the temperature rise response time caused by a load disturbance includes: acquiring the temperature rise response time range and determining candidate temperature rise response times; matching the time interval between the candidate temperature rise response times and the time when the load disturbance occurs with a preset thermal response time range, and simultaneously determining whether the current change amplitude of the load disturbance and the temperature rise change amount satisfy a positive correlation; if satisfied, confirming that the temperature rise change is caused by the load disturbance; using the confirmed candidate temperature rise response times as a benchmark, searching along the temperature sampling time sequence direction for the first sampling time that simultaneously satisfies all the determination conditions of temperature rise change amount, temperature rise change rate, and temperature rise duration, and correcting it as the temperature rise response time; performing consistency verification on the temperature change data within a preset range after the corrected temperature rise response time, and if the verification passes, outputting the time as the final temperature rise response time, otherwise continuing to search along the time sequence until the verification passes.

[0011] In one embodiment, obtaining the temperature rise response time range includes: calculating disturbance heat characterization parameters based on current change data before and after the load disturbance occurs; matching historical load disturbance samples with the disturbance heat characterization parameters to obtain an initial temperature rise response time range; identifying the thermal inertia parameters of the power metering box based on temperature change data within the initial temperature rise response time range; determining the heat influence time range based on the disturbance heat characterization parameters, and determining the thermal inertia influence time range based on the thermal inertia parameters; and fusing the heat influence time range and the thermal inertia influence time range to obtain the dynamic temperature rise response time range.

[0012] In one embodiment, constructing a dynamic delay feature between current disturbance and temperature rise response includes: obtaining a base delay time based on the time difference between the occurrence of the load disturbance and the temperature rise response; establishing a dynamic delay mapping relationship between the current disturbance feature and the temperature rise response delay time based on the dynamic delay feature data of historical load disturbance events; inputting the current disturbance feature corresponding to the current load disturbance into the dynamic delay mapping relationship to obtain the theoretical delay time corresponding to the current load disturbance; comparing the base delay time with the theoretical delay time, and confirming the validity of the current dynamic delay feature when the deviation between the two is within a preset allowable range, and using the confirmed valid dynamic delay feature for identifying abnormal heating status of the power metering box.

[0013] In one embodiment, establishing a dynamic delay mapping relationship between current disturbance characteristics and temperature rise response delay time includes: establishing an initial mapping model between current disturbance characteristic parameters and temperature rise response delay time based on a set of historical dynamic delay characteristics; obtaining the deviation between the current thermal state parameters of the equipment and the standard thermal state parameters under normal operating conditions, and determining the thermal state correction coefficient based on the deviation; and using the thermal state correction coefficient to correct the initial mapping model to obtain a dynamic delay mapping relationship after adaptive correction of the equipment thermal state.

[0014] In one embodiment, the dynamic delay characteristics are matched with the thermal response characteristics corresponding to the normal operation of the power metering box. Based on the matching results, it is determined whether the power metering box has an abnormal heating fault. This includes: obtaining the real-time thermal response characteristic parameters corresponding to the current load disturbance and comparing them with the normal thermal response characteristics to obtain the deviation of each thermal response characteristic parameter; weighting and fusing each deviation to obtain a comprehensive thermal response deviation index; when the comprehensive thermal response deviation index does not exceed the preset abnormal judgment threshold, it is determined that there is no abnormal heating fault; otherwise, it is determined that there is an abnormal heating risk, and the fault type is assisted in judging based on the degree of abnormality of the dynamic delay characteristics and the temperature rise change characteristics, and the abnormal heating judgment result is output.

[0015] In one embodiment, fault type determination is aided by analyzing the degree of anomaly in dynamic delay characteristics and temperature rise change characteristics. This includes: obtaining the deviation between the actual temperature rise response delay time corresponding to the current load disturbance and the standard temperature rise response delay time under normal operating conditions, as the degree of dynamic delay anomaly; obtaining the deviation between the temperature rise change characteristics during the current temperature rise response and the corresponding standard value under normal operating conditions, as the degree of temperature rise anomaly; and determining the abnormal heating type based on the combination relationship between the degree of dynamic delay anomaly and the degree of temperature rise anomaly: when the dynamic delay is abnormal and the temperature rise amplitude is normal, it is determined to be an abnormal heat transfer or thermal inertia; when the temperature rise amplitude is abnormal and the dynamic delay is normal, it is determined to be a local abnormal heating; when the temperature rise amplitude is abnormal and the temperature rise rate is abnormal, it is determined to be a rapid heat accumulation anomaly; and when the dynamic delay is abnormal and the temperature rise duration is abnormal, it is determined to be an abnormal decrease in heat dissipation capacity.

[0016] Secondly, this application provides an IoT-based fault diagnosis system for power metering boxes. This system includes: a load disturbance identification module, used to identify load disturbance events based on current change characteristics during the operation of the power metering box and determine the time of occurrence of the load disturbance; a temperature rise response analysis module, used to analyze the temperature response process within a preset time range based on the time of occurrence of the load disturbance and determine the time of temperature rise response caused by the load disturbance; a dynamic delay feature construction module, used to construct dynamic delay features between current disturbance and temperature rise response based on the time difference between the time of occurrence of the load disturbance and the time of temperature rise response; and an abnormal heating diagnosis module, used to match the dynamic delay features with the corresponding thermal response features under normal operating conditions of the power metering box, and determine whether the power metering box has an abnormal heating fault based on the matching results.

[0017] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: By collecting current change and temperature response information during the operation of the power metering box, a dynamic correlation is established between load disturbance, electrothermal transfer process, and temperature rise response, realizing a full-process state analysis from the source of current disturbance to the result of thermal response. Compared with traditional methods that rely solely on temperature thresholds or single temperature rise amplitude for fault detection, this solution utilizes the dynamic delay characteristics between the occurrence of load disturbance and the timing of temperature rise response to effectively characterize heat transfer, changes in thermal inertia, and thermal response deviations caused by abnormal heating within the power metering box, improving the accuracy and reliability of abnormal heating fault detection. Simultaneously, by matching the real-time dynamic delay characteristics with the thermal response characteristics under normal operating conditions, the impact of ambient temperature changes, load fluctuations, and differences in equipment operating conditions on the detection results can be reduced, enabling accurate identification of early abnormal heating states, thereby improving the operational safety and intelligent fault diagnosis level of the power metering box. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the fault diagnosis method for an IoT-based power metering box provided in an embodiment of this application.

[0020] Figure 2 This is a schematic diagram of the structure of an IoT-based power metering box fault diagnosis system provided in an embodiment of this application.

[0021] Figure 3 This is a schematic diagram illustrating the relationship between current disturbance and temperature rise response time delay, provided for an embodiment of this application.

[0022] Figure 4 A comparison chart of dynamic delay deviation index between normal state and abnormal heating state provided in the embodiments of this application. Detailed Implementation

[0023] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Reference Figure 1As shown in the schematic diagram, the fault diagnosis method for an IoT-based power metering box provided by this invention includes the following steps: S1 identifies load disturbance events based on the current change characteristics during the operation of the power metering box and determines the time when the load disturbance occurs.

[0025] In this embodiment, load disturbance events are identified based on the current change characteristics during the operation of the power metering box, and the time of occurrence of the load disturbance is determined, including: S11, collect continuous current monitoring data during the operation of the power metering box, and establish a current change time series dataset; Specifically, by using a current detection device installed inside or outside the power metering box, and an IoT data acquisition and communication module, the operating current of the power metering box during operation is continuously sampled. The current value corresponding to each sampling moment is obtained according to a preset sampling period, and the current values ​​are arranged according to the sampling time sequence to form a current change time series dataset containing time information and current amplitude information. The current change time series dataset is used to characterize the dynamic current change process of the power metering box under normal operation and load change conditions.

[0026] S12, Based on the current change time series dataset, calculate the current change between adjacent sampling times to obtain the current disturbance change characteristics; Specifically, based on the current change time series dataset, the current difference between the current sampling time and the previous sampling time is calculated to obtain the current change sequence. Further, the current change sequence is normalized using the max-min normalization method to eliminate the influence of current amplitude differences under different load operating conditions on the disturbance identification results. Current disturbance change features used to characterize the degree of load change are extracted based on the normalized current change. Among them, the current disturbance change features include at least the current change amplitude, the current change rate, and the duration of change.

[0027] S13, Match the current disturbance change characteristics with the preset disturbance judgment conditions to identify load disturbance events during the operation of the power metering box; Specifically, the characteristics of current disturbance changes are compared with pre-established load disturbance judgment thresholds. When the amplitude of current change exceeds the preset change threshold, the rate of current change exceeds the preset rate threshold, and the duration of current change meets the preset duration condition, the current current change process is determined to be a load disturbance event. Otherwise, the current current change is determined to be a normal load fluctuation and is not recorded as a load disturbance event.

[0028] The preset change threshold is determined by collecting historical current data from the electricity metering box under normal operating conditions, statistically analyzing the maximum current change amplitude during normal load fluctuations, and using a preset proportion greater than the maximum current change amplitude as the preset change threshold, for example, 1.1 to 1.5. When the proportion is less than 1.1, the threshold is too close to the normal fluctuation range, which may easily lead to false triggering due to measurement errors, sampling noise, or environmental changes; when the proportion is greater than 1.5, the judgment condition is too strict, which may lead to the failure to identify small load disturbance events with heat accumulation effects in a timely manner, thus distinguishing between normal load fluctuations and load disturbance events. The preset rate threshold is determined by statistically analyzing the current change rate at adjacent sampling times under normal operating conditions of the electricity metering box, obtaining the maximum current change rate during normal operation, and using a preset proportion greater than the maximum current change rate as the preset rate threshold. The preset duration is determined by statistically analyzing the time range of current change in historical load disturbance event samples, determining the minimum duration that can reflect the actual load switching or load change process, and setting this minimum duration as the preset duration.

[0029] S14. Determine the candidate occurrence time of the load disturbance based on the starting position of the current change corresponding to the load disturbance event; Specifically, after identifying the load disturbance event, based on the current change time series dataset, the continuous current change interval that meets the load disturbance judgment condition is obtained, and the starting sampling time of the continuous change interval is determined; the starting sampling time is used as the candidate occurrence time of the load disturbance to characterize the time node when the power metering box enters the disturbance change state from the original stable load state.

[0030] S15, perform time correction on the candidate occurrence time of load disturbance to determine the occurrence time of load disturbance.

[0031] Furthermore, time correction is performed on the candidate occurrence times of load disturbances to determine the occurrence time of the load disturbance, including: S151, determine the disturbance correction time window based on the candidate occurrence time of the load disturbance, and acquire the current sampling data within the disturbance correction time window; Specifically, the candidate occurrence time of the load disturbance obtained in step S14 is used as the time reference point. A disturbance correction time window is set according to the current sampling period of the power metering box and the load change response time. The disturbance correction time window includes a pre-disturbance analysis interval before the candidate occurrence time of the load disturbance and a post-disturbance analysis interval after the candidate occurrence time of the load disturbance. Further, the current values ​​corresponding to each sampling time within the disturbance correction time window are extracted from the current change time series dataset to form a local current change sequence for time correction.

[0032] The length of the disturbance correction time window is determined based on the duration of load changes in the historical operating data of the power metering box, and the time range that can cover a complete load change process is used as the disturbance correction time window.

[0033] S152, calculate the characteristics of the steady current before the disturbance based on the local current change sequence, and determine the steady state before the disturbance; Specifically, based on the current sampling data within the analysis interval before the disturbance, the average current value of all sampling points within that interval is obtained as the reference current value before the disturbance; at the same time, the deviation between the current value of each sampling point and the reference current value before the disturbance is obtained, and the current fluctuation range before the disturbance is determined based on the deviation.

[0034] When the current fluctuation range within the analysis interval before the disturbance is less than the preset stable fluctuation threshold, it is determined that the power metering box was in a stable operating state before the candidate time of the load disturbance, and the corresponding interval is taken as the stable state interval before the disturbance.

[0035] The preset stable fluctuation threshold is determined by collecting historical current sampling data under normal operating conditions of the power metering box, obtaining the difference between the maximum and minimum current values ​​within each continuous stable operating time window, sorting the current fluctuation ranges corresponding to multiple time windows, and determining the maximum current fluctuation range in the top 95% as the preset stable fluctuation threshold.

[0036] S153, determine the current change benchmark based on the steady state interval before the disturbance, and extract the current change characteristics after the candidate occurrence time of the load disturbance; Specifically, based on the reference current value before the disturbance, the change in current at each sampling moment within the analysis interval after the disturbance is calculated, where the change in current is the difference between the current value at the current sampling moment and the reference current value before the disturbance. Furthermore, the rate of change of current is extracted based on the amount of change of current between adjacent sampling times, and the number of sampling points that continuously meet the current change conditions is counted to obtain current change characteristics including the amplitude of current change after disturbance, the rate of change of current, and the duration of current. The amplitude of current change is used to characterize the degree of change in current magnitude before and after load change; the rate of change of current is used to characterize the speed at which load change occurs; and the duration of current is used to characterize whether the current change is a continuous load change rather than an instantaneous disturbance.

[0037] S154, determine whether the candidate occurrence time of load disturbance is the actual start time of disturbance based on the characteristics of current change; Specifically, the current change amplitude, current change rate, and current duration are compared with preset change thresholds, preset rate thresholds, and preset durations, respectively. When the current change amplitude is greater than or equal to a preset change threshold, and the current change rate is greater than or equal to a preset rate threshold, and the duration of time that continuously meets the above conditions is greater than or equal to a preset duration, then the candidate occurrence time of the load disturbance is determined to correspond to the actual load disturbance process, and the candidate occurrence time of the load disturbance is retained.

[0038] If any condition is not met, it is determined that there is a time offset in the time when the load disturbance candidate occurs.

[0039] S155, when there is a time offset in the candidate occurrence time of load disturbance, the corrected occurrence time of disturbance is determined according to the first sampling time that continuously meets the disturbance judgment condition; Specifically, starting from the sampling point corresponding to the candidate occurrence time of load disturbance, the current change characteristics corresponding to each sampling time are searched sequentially according to the current sampling time order. For each sampling time, extract the current change amplitude relative to the reference current value before the disturbance, the change rate between adjacent sampling points, and determine whether the consecutive sampling points meet the preset change threshold, preset rate threshold, and preset duration requirements. The sampling time that first satisfies all disturbance determination conditions is determined as the corrected load disturbance occurrence time.

[0040] S156, perform consistency verification on the time of load disturbance occurrence, and output the final time of load disturbance occurrence; Specifically, using the corrected time of the load disturbance as the time reference, the current change data within a preset verification time range after that time is reacquired, and it is determined whether the current change data continues to meet the load disturbance characteristics. If the current change after the corrected load disturbance still meets the requirements for disturbance change amplitude, change rate and duration, then the corrected time point is determined as the final load disturbance occurrence time. If the conditions are not met, return to step S155 to search for the next sampling time that meets the conditions, until the final load disturbance time is determined.

[0041] It should be noted that by collecting continuous current monitoring data during the operation of the power metering box, a current change time series dataset is constructed. This dataset, combined with multi-dimensional disturbance characteristics such as current change amplitude, current change rate, and duration, identifies load disturbance events. Compared to methods relying solely on a single current surge value, this approach effectively reduces misjudgments caused by normal load fluctuations, transient interference, and sampling noise. Furthermore, by determining preset change thresholds, preset rate thresholds, and preset durations based on historical operating data, the disturbance judgment parameters are adapted to the actual operating state of the power metering box, improving the reliability of load disturbance identification. After determining the candidate occurrence time of the load disturbance, this scheme further utilizes a disturbance correction time window to analyze the changes in current state before and after the disturbance. By establishing a stable current benchmark before the disturbance and comprehensively judging the current change amplitude, change rate, and duration after the candidate time, accurate correction of the disturbance start time is achieved. When there is a time offset in the candidate time, the sampling time that first meets the disturbance conditions is searched along the current time series direction, and a consistency verification mechanism is used to determine the final load disturbance occurrence time, thereby avoiding time positioning deviations caused by current acquisition delays, transient impacts, or random fluctuations. Therefore, this solution can more accurately obtain the time of load disturbance, providing a reliable time reference for subsequent analysis of the temperature rise response process caused by load disturbance, construction of dynamic delay characteristics between current disturbance and thermal response, and judgment of abnormal heating faults in the power metering box. It has the advantages of high identification accuracy, strong anti-interference ability and low implementation cost.

[0042] S2, based on the time of load disturbance, analyze the temperature response process within a preset time range to determine the time of temperature rise response caused by load disturbance.

[0043] In this embodiment, based on the moment of load disturbance, the temperature response process within a preset time range is analyzed to determine the moment of temperature rise response caused by the load disturbance, including: S21, determine the temperature rise response time range based on the time of load disturbance occurrence, and obtain the temperature change time series dataset within the corresponding time range; Specifically, based on the temperature rise response time range, temperature detection data at each sampling time within the corresponding time period are obtained from temperature detection devices located inside or outside the power metering box, and arranged in the order of sampling time to form a time series dataset of temperature changes corresponding to the time when the load disturbance occurs.

[0044] S22, extract temperature change features before and after load disturbance based on temperature change time series dataset; Specifically, based on the temperature change time series dataset, the temperature data within a preset time period before the load disturbance occurs is obtained, and the average temperature value and temperature fluctuation range within this time period are extracted to obtain the temperature baseline state before the disturbance. Furthermore, the temperature values ​​corresponding to each sampling time after the occurrence of the load disturbance are obtained, and the temperature change of each sampling time relative to the temperature reference state before the disturbance is obtained, so as to obtain the temperature rise change characteristics. Among them, the temperature rise change is used to characterize the degree of increase in the internal temperature of the power metering box relative to the original steady state after the load disturbance occurs.

[0045] S23, Identify the temperature rise response process based on the temperature change characteristics and determine the candidate time of the temperature rise response; Specifically, based on the temperature change characteristics, each temperature sampling point after the occurrence of the load disturbance is analyzed point by point to obtain the temperature rise change and the temperature rise rate corresponding to each sampling time. If the temperature change after a certain sampling time continuously reaches a preset temperature rise threshold, and the corresponding temperature change rate continuously reaches a preset temperature rise rate threshold, then the temperature data corresponding to that sampling time is determined to have entered a temperature rise change state.

[0046] The sampling time that first meets the above temperature rise determination conditions is taken as the candidate time of temperature rise response, which is used to characterize the time node when the temperature changes from a steady state to a rising state. Among them, the preset temperature rise change threshold is determined based on the temperature fluctuation range under normal operating conditions of the power metering box, and the temperature rise amplitude that exceeds the normal temperature fluctuation range is used as the preset temperature rise change threshold; the preset temperature rise rate threshold is determined based on the statistical results of the temperature change rate under normal operating conditions, and the temperature rise rate that exceeds the normal temperature change rate is used as the preset temperature rise rate threshold.

[0047] S24. Based on the correlation between the candidate time of temperature rise response and the time of occurrence of load disturbance, determine whether the temperature rise change is caused by load disturbance; Specifically, based on the load disturbance occurrence time determined in step S1 and the candidate temperature rise response time determined in step S23, the time interval between the two is calculated:

[0048] In the formula, The time interval between load disturbance and temperature rise response. Candidate moments for temperature rise response The time when the load disturbance occurs.

[0049] Match the time interval with the preset thermal response time range; When the time interval is within the preset thermal response time range, it indicates that the temperature rise change conforms to the time characteristics of the heat transfer process after load disturbance; The preset thermal response time range is determined by collecting load disturbance data and corresponding temperature response data under normal operating conditions of the power metering box, calculating the time difference between the time of load disturbance occurrence and the time of temperature rise response within different disturbance heat ranges, and sorting them from smallest to largest according to the time difference. The time range covering 90% to 95% of historical samples is then determined as the preset thermal response time range.

[0050] Furthermore, a correlation analysis is performed based on the current change characteristics and temperature rise change characteristics corresponding to the load disturbance; Obtain the current change amplitude during load disturbance and determine whether there is a positive correlation between the current change amplitude and the temperature rise change. Specifically, the correlation coefficient between the current change amplitude and the temperature rise change corresponding to the load disturbance is obtained, and the correlation coefficient is compared with a preset correlation threshold. When the correlation coefficient is greater than or equal to the preset correlation threshold, it is determined that the current change amplitude and the temperature rise change satisfy a positive correlation relationship. The correlation coefficient is obtained by collecting data on the amplitude of current change and the amount of temperature rise corresponding to multiple historical load disturbance events, and the Pearson correlation coefficient calculation method is used. The preset correlation threshold is determined by statistically analyzing the correlation coefficient distribution of historical load disturbance events under normal operating conditions, and the lowest correlation coefficient value covering 90% to 95% of normal load disturbance samples is determined as the preset correlation threshold.

[0051] When the amplitude of the current change increases, the corresponding temperature rise change increases synchronously. If the temperature rise change meets the preset thermal response time range, it is determined that the temperature rise change is caused by the load disturbance. Otherwise, it is determined that the temperature rise change may be caused by changes in ambient temperature, external heat source influence, or other non-load factors, and the corresponding candidate temperature rise response time is eliminated.

[0052] S25, perform time correction on the confirmed candidate time of temperature rise response, and determine the time of temperature rise response; Specifically, when step S24 determines that the temperature rise change is caused by load disturbance, the candidate time of temperature rise response is used as the initial temperature rise response time; Furthermore, a temperature rise correction time window is established with the initial temperature rise response time as the center, and the temperature rise change and temperature rise rate corresponding to each temperature sampling point are re-analyzed within the temperature rise correction time window. For each temperature sampling moment, determine whether it meets the following conditions for the first time: the temperature rise change reaches the preset temperature rise change threshold, the temperature rise change rate reaches the preset temperature rise rate threshold, and the duration reaches the preset temperature rise duration. The temperature sampling moment that first meets the above conditions is determined as the corrected temperature rise response moment.

[0053] The preset temperature rise duration is calculated by collecting temperature change data under normal operation and historical load disturbance conditions of the power metering box, statistically analyzing the duration of temperature rise changes, and setting the minimum duration that can distinguish between transient temperature fluctuations and continuous temperature rise processes as the preset temperature rise duration.

[0054] S26, verify the validity of the temperature rise response time and output the final temperature rise response time; Specifically, the temperature rise response time determined in step S25 is used as the time reference to obtain temperature change data within a preset verification range after that time. Furthermore, determine whether the temperature change after the temperature rise response time continues to satisfy the temperature rise response characteristics; If the temperature continues to rise and the temperature rise amplitude reaches the preset amplitude requirement, the temperature rise response time is determined to be valid and is taken as the final temperature rise response time. If the temperature does not continue to rise or the temperature rise amplitude does not reach the preset amplitude requirement, it is determined that the current temperature rise response time is misjudged, and the process returns to step S25 to search for a temperature rise response time that meets the conditions.

[0055] Specifically, by statistically analyzing the temperature rise amplitude range corresponding to the temperature rise stabilization phase in historical load disturbance events, the lowest temperature rise amplitude value that can characterize the effective temperature rise response process is determined as the preset amplitude requirement.

[0056] Furthermore, to verify the existence of a thermal inertia delay relationship between load disturbance and temperature rise response, current and temperature change data during the same load disturbance process were analyzed simultaneously to obtain current disturbance and temperature rise response curves, as shown below. Figure 3 As shown.

[0057] like Figure 3 As shown, when the current suddenly changes at 60s, the internal temperature of the metering chamber does not immediately change synchronously. Instead, it begins to show a significant upward trend after approximately 18s of heat transfer time, and reaches a stable temperature rise state around 90s. Therefore, by analyzing the time difference between the occurrence time of the load disturbance and the start time of the temperature rise response, the timing of the temperature rise response caused by the load disturbance can be effectively extracted, avoiding misjudging ambient temperature fluctuations or transient temperature changes as abnormal heating.

[0058] Furthermore, determining the temperature rise response time range based on the timing of the load disturbance also includes: S211, Obtain the corresponding current disturbance data based on the time of load disturbance occurrence, and calculate the heat characterization parameters generated by the load disturbance; Specifically, using the time of load disturbance occurrence determined in step S1 as the time reference point, current monitoring data are obtained within a preset time range before the time of load disturbance occurrence and during the duration of load disturbance thereafter. Based on the stable operating current value before the load disturbance and the real-time current value during the load disturbance, calculate the current disturbance amount corresponding to each sampling time.

[0059] In the formula, Let be the current disturbance at the i-th sampling time. This represents the actual operating current value at the i-th sampling time. This is the stable operating current value before the load disturbance occurs.

[0060] Furthermore, based on the current disturbance magnitude and duration, the degree of heat accumulation during the load disturbance process is calculated to obtain disturbance heat characterization parameters:

[0061] In the formula, The parameter characterizing the heat of disturbance generated by load disturbance. These are the equivalent resistance parameters of the conductive components corresponding to the power metering box. For the current sampling period, This represents the number of sampling points during the duration of the load disturbance.

[0062] S212, determine the initial temperature rise response time range based on the heat characterization parameters; Specifically, a historical load disturbance sample database is established, which includes the disturbance heat characterization parameters corresponding to the historical load disturbance, the corresponding temperature rise response start time, and the corresponding temperature rise duration. Furthermore, based on the heat characterization parameters corresponding to the current load disturbance, the temperature rise response duration corresponding to the current heat characterization parameters is matched in the historical load disturbance sample database, and the matched time length is used as the initial temperature rise response time range; S213, Identify the thermal inertia parameters of the power metering box based on the initial temperature rise response time range; Specifically, temperature change data after load disturbance is obtained based on the initial temperature rise response time range, and a thermal response model of the power metering box is established, expressed as:

[0063] In the formula, Let be the temperature detected at time t. The initial temperature before the disturbance occurred. To stabilize the temperature rise, These are the thermal inertia parameters of the electricity metering box.

[0064] Based on continuous temperature sampling data after load disturbance, the thermal response model is solved by least squares fitting method to obtain thermal inertia parameters; Among them, the thermal inertia parameter is used to characterize the time required for the power metering box to reach a steady temperature after heat input.

[0065] S214, the initial temperature rise response time range is corrected according to the thermal inertia parameters to obtain the dynamic temperature rise response time range; Specifically, the time range of heat influence is determined based on the disturbance heat characterization parameters:

[0066] In the formula, For the time range of heat effect, , This is a proportionality coefficient determined based on historical load disturbance samples.

[0067] The proportional coefficient is obtained by collecting temperature rise response time data corresponding to different disturbance heat characterization parameters under normal operation of the power metering box, obtaining the proportional relationship between heat influence time and disturbance heat characterization parameters under each disturbance heat condition, and statistically analyzing the proportional values ​​corresponding to multiple historical samples, and using the statistically obtained average proportional value as the proportional coefficient.

[0068] The time range of thermal inertia influence is determined based on the thermal inertia parameter τ.

[0069] In the formula, This represents the temperature rise response time range corresponding to thermal inertia. This is the preset response scaling factor.

[0070] The preset response ratio coefficient is obtained by collecting data on the thermal inertia parameters and the time required for the temperature rise to reach a stable response state under normal operating conditions of the power metering box. The ratio relationship between the thermal inertia parameters and the temperature rise response time in different samples is obtained, and the ratio range covering 90% to 95% of historical normal samples is determined as the preset response ratio coefficient.

[0071] By combining the time ranges of heat effect and thermal inertia effect, the dynamic temperature rise response time range is obtained:

[0072] In the formula, This refers to the dynamic temperature rise response time range.

[0073] It should be noted that the temperature response analysis process is established based on the moment of load disturbance. By acquiring time-series data of temperature changes before and after the disturbance, features such as the amount and rate of temperature rise are extracted to accurately identify the transition from a stable to a rising temperature state. Simultaneously, by introducing heat characterization parameters corresponding to the load disturbance and thermal inertia parameters of the power metering box, the temperature rise response time range is dynamically determined. This ensures that the temperature rise analysis range no longer relies on fixed empirical values ​​but can be adaptively adjusted according to different load disturbance intensities and the thermal response characteristics of different equipment. This avoids the problem of missing the temperature rise response process due to an excessively short time range, or the interference of invalid data such as ambient temperature changes due to an excessively long time range. Furthermore, this scheme matches the time interval between the candidate temperature rise response moment and the moment of load disturbance occurrence with thermal response patterns, and combines the correlation between the current disturbance amplitude and the amount of temperature rise change to determine the source of temperature rise change. This effectively eliminates temperature changes caused by non-load factors such as ambient temperature changes and external heat source influences, improving the accuracy of temperature rise response identification. Furthermore, by performing time correction and validity verification on the candidate temperature rise response times, the time offset caused by temperature sensor sampling period, heat transfer delay, and data processing errors is reduced. This makes the final determined temperature rise response time closer to the start time of the thermal response caused by the actual load disturbance, providing a more accurate and reliable time reference for subsequently constructing the dynamic delay characteristics between current disturbance and temperature rise response, and for judging abnormal heating faults in the power metering box. Overall, this technical solution establishes an electrothermal coupling analysis mechanism based on current disturbance-heat generation-thermal inertia transfer-temperature rise response, which has the advantages of high accuracy in determining the temperature rise response time, strong adaptability, strong resistance to environmental interference, and applicability to power metering boxes under different operating conditions.

[0074] S3. Based on the time difference between the occurrence of load disturbance and the temperature rise response, construct the dynamic delay characteristics between current disturbance and temperature rise response.

[0075] In this embodiment, based on the time difference between the occurrence of the load disturbance and the temperature rise response, a dynamic delay characteristic between the current disturbance and the temperature rise response is constructed, including: S31, determine the basic delay time from current disturbance to temperature rise response based on the time of load disturbance occurrence and temperature rise response; Specifically, based on the load disturbance occurrence time determined in step S1 and the temperature rise response time determined in step S2, the corresponding time stamps are obtained, and the time difference between the load disturbance occurrence time and the temperature rise response time is calculated to obtain the basic delay time between the current disturbance and the temperature rise response.

[0076] The base delay time is used to reflect the combined time delay caused by heat transfer, heat diffusion, and temperature detection response inside the power metering box under the current load disturbance.

[0077] S32, based on the current change characteristics corresponding to the load disturbance, the basic delay time is correlated and described to obtain the dynamic delay characteristics of a single disturbance; Specifically, based on step S1, the current change characteristics corresponding to the load disturbance event are determined, and the current change amplitude, current change rate and load disturbance duration corresponding to the load disturbance are obtained. By associating the current change characteristics with the baseline delay time obtained in step S31, a dynamic delay feature vector corresponding to the current load disturbance is constructed:

[0078] In the formula, The dynamic delay characteristic corresponding to the i-th load disturbance is... Let be the base delay time from the current disturbance to the temperature rise response corresponding to the i-th load disturbance. Let be the amplitude of the current change corresponding to the i-th load disturbance. Let be the rate of change of current corresponding to the i-th load disturbance. The duration of the load disturbance corresponding to the i-th load disturbance.

[0079] Through the above steps, the single time difference parameter is extended into a dynamic delay feature that includes load disturbance state information, so that the dynamic delay feature can simultaneously reflect the load disturbance intensity and the corresponding thermal response hysteresis process.

[0080] S33, statistically analyze the dynamic delay characteristics corresponding to multiple historical load disturbance events, and establish a dynamic delay characteristic set; Specifically, steps S31 to S32 are repeated to obtain the dynamic delay feature vectors corresponding to multiple confirmed load disturbance events during the historical operation of the power metering box, and a dynamic delay feature set is established based on the dynamic delay feature vectors corresponding to each historical load disturbance event. Furthermore, statistical analysis is performed on the delay time parameter, current change amplitude parameter, current change rate parameter, and load disturbance duration parameter in the dynamic delay feature set to obtain the dynamic delay variation law between current disturbance and temperature rise response under normal operation of the power metering box. The statistical analysis includes extracting the average value and variation range of the delay time parameter, and analyzing the corresponding delay variation trend under different current change amplitude, different current change rate, and different load disturbance duration conditions.

[0081] S34, construct the dynamic delay mapping relationship between current disturbance and temperature rise response based on the dynamic delay feature set; S35, determine the real-time dynamic delay parameters based on the dynamic delay characteristics corresponding to the current load disturbance; Specifically, when a new load disturbance event is identified in step S1 and the corresponding temperature rise response time is determined in step S2, the dynamic delay mapping relationship established in step S34 is input based on the current change amplitude, current change rate and load duration corresponding to the current load disturbance. Further, the theoretical temperature rise response delay time corresponding to the current load disturbance is obtained, and the theoretical temperature rise response delay time is compared with the actual base delay time calculated in step S31; When the deviation between the actual base delay time and the theoretical temperature rise response delay time is within the preset allowable range, the dynamic delay characteristic corresponding to the current load disturbance is determined to be valid.

[0082] S36, normalize the dynamic delay features to form dynamic delay feature parameters for judging abnormal heating; Specifically, since the basic delay time, current change amplitude, current change rate, and load disturbance duration corresponding to different load disturbance events have different dimensions, the basic delay time, current change amplitude, current change rate, and load disturbance duration in the dynamic delay feature set are normalized using the maximum-minimum normalization method to obtain dimensionless dynamic delay feature parameters. Based on the normalized basic delay time parameters, current change amplitude parameters, current change rate parameters, and load disturbance duration parameters, the final dynamic delay features for identifying abnormal heating states are formed.

[0083] S37 outputs dynamic delay features for identifying abnormal heating status of the power metering box; Specifically, the dynamic delay characteristic parameters obtained in step S36 are used as characteristic data to characterize the dynamic coupling relationship between current disturbance and temperature rise response. The dynamic delay characteristics include at least: the basic time delay between the occurrence of load disturbance and the occurrence of temperature rise response, current disturbance amplitude, current change rate, load disturbance duration, temperature rise response delay variation law corresponding to different current disturbance amplitudes, thermal response delay characteristics corresponding to different current change rates, thermal response delay variation law corresponding to different load disturbance durations, and the deviation between the actual delay time and the normal delay law under the current disturbance state.

[0084] Specifically, to further illustrate that dynamic delay characteristics can reflect the differences in thermal response of the electricity metering box under different load disturbance conditions, multiple historical load disturbance events were selected for statistical analysis to obtain dynamic delay characteristic parameters corresponding to different disturbance conditions, as shown in Table 1.

[0085] As shown in Table 1, the temperature rise response delay time exhibits a certain variation pattern as the current disturbance amplitude, disturbance duration, and disturbance heat increase. When the equipment is in a normal thermal state, the dynamic delay parameters under different disturbance conditions remain within a reasonable range; when the thermal state of the equipment changes, the dynamic delay time will shift, thus serving as an important basis for identifying abnormal heating states.

[0086] Table 1 Dynamic delay characteristic parameters under different load disturbances

[0087] Further, S34, a dynamic delay mapping relationship between current disturbance and temperature rise response is constructed based on the dynamic delay feature set, including: S341, extract current disturbance characteristic parameters, temperature rise response basic delay parameters and corresponding equipment thermal state parameters based on the dynamic delay characteristic set; The equipment thermal status parameters include the stable temperature value before disturbance, temperature change trend, historical temperature rise recovery time, continuous operation time, and ambient temperature parameters.

[0088] S342, Determine the thermal state correction coefficient based on the equipment thermal state parameters; Specifically, based on the equipment's thermal status parameters, the current equipment thermal status parameters are compared with the standard thermal status parameters corresponding to the normal operating state of the power metering box; The standard thermal state parameters under normal operating conditions are obtained by collecting historical operating data of the power metering box under conditions where there is no abnormal heating.

[0089] Calculate the deviation between the current thermal state and the normal thermal state of the equipment:

[0090] In the formula, This refers to the deviation in the thermal condition of the equipment. This is the thermal state reference value corresponding to normal operating conditions. This corresponds to the current thermal state value.

[0091] Furthermore, the thermal state correction factor is calculated based on the thermal state deviation:

[0092] In the formula, This is a thermal state correction factor. This is the temperature reference offset. This is the temperature recovery time offset. This is the continuous runtime offset. This represents the offset of the temperature change trend. This represents the offset of the ambient temperature parameter. , , , , These are the corresponding weighting coefficients.

[0093] The weighting coefficients are determined by collecting historical thermal state parameter data under normal operating conditions of the power metering box, calculating the correlation between the change in each thermal state parameter and the change in temperature rise response delay time, and normalizing the coefficients according to the proportion of the correlation of each parameter to the total correlation of all parameters.

[0094] S343, Establish an initial dynamic delay mapping model for current disturbance and temperature rise response based on the dynamic delay feature set; Specifically, the initial dynamic delay mapping model is as follows:

[0095] In the formula, This is the theoretical temperature rise response delay time. The magnitude of the current change. The rate of change of current, The duration of the disturbance. The parameters characterizing the heat of disturbance, This represents the dynamic mapping relationship between the current disturbance characteristic parameters and the temperature rise response delay time.

[0096] Furthermore, based on the historical dynamic delay feature set, the initial dynamic delay mapping model is fitted with parameters to obtain the corresponding model parameters:

[0097] In the formula, , , , These are the influence coefficients corresponding to each disturbance characteristic parameter. These are the model bias parameters.

[0098] The influence coefficient is determined by collecting historical dynamic delay feature sets, calculating the correlation between the change in each disturbance feature parameter and the change in temperature rise response delay time, and normalizing it according to the proportion of the correlation of each disturbance feature parameter to the total correlation of all disturbance feature parameters.

[0099] Through the above steps, a basic mapping relationship between current disturbance characteristics and temperature rise response delay time is established.

[0100] S344, The initial dynamic delay mapping model is corrected according to the thermal state correction coefficient to obtain the dynamic delay mapping relationship; Specifically, the thermal state correction coefficient obtained in step S342 is introduced into the initial dynamic delay mapping model to correct the basic delay time;

[0101] In the formula, This is the dynamic delay time after the equipment's thermal condition has been corrected.

[0102] It should be noted that by associating the time difference between the occurrence of load disturbance and the temperature rise response with the amplitude of current change, the rate of current change, the duration of disturbance, the heat of disturbance, and the thermal state parameters of the equipment, a dynamic delay mapping relationship including equipment thermal state feedback correction is constructed. Compared with the method that only uses a fixed time difference as the judgment basis, this method can more comprehensively characterize the dynamic coupling relationship between current disturbance and temperature rise response caused by the degree of load change, heat transfer process, and the thermal inertia of the equipment itself. At the same time, by establishing an initial mapping model based on the historical dynamic delay feature set and using the current equipment thermal state parameters to adaptively correct the model, the obtained dynamic delay features can adapt to the changes in response characteristics of the power metering box caused by long-term operation, environmental changes, changes in heat dissipation capacity, and thermal performance drift, thus improving the accuracy and stability of temperature rise response delay prediction under different operating conditions. In addition, by comparing the actual delay time with the model predicted delay time, this scheme can effectively identify abnormal delay phenomena that deviate from the normal thermal response law, providing more accurate and reliable feature basis for subsequent abnormal heating fault detection of the power metering box.

[0103] S4. Match the dynamic delay characteristics with the thermal response characteristics corresponding to the normal operation of the power meter box, and determine whether there is an abnormal heating fault in the power meter box based on the matching results.

[0104] In this embodiment, the dynamic delay characteristics are matched with the thermal response characteristics corresponding to the normal operating state of the power metering box, and the abnormal heating fault of the power metering box is determined based on the matching result, including: S41, acquire the thermal response characteristics corresponding to the normal operation of the power metering box, and establish a normal thermal response characteristic library; Specifically, based on the historical operating data of the power metering box under normal operating conditions, load disturbance event data, temperature change data, and thermal response process data are obtained when there is no abnormal heating fault. Among them, the thermal response characteristics under normal operating conditions include the temperature rise response delay time after load disturbance, temperature rise change amplitude, temperature rise change rate, temperature rise duration, and temperature rise recovery time.

[0105] Furthermore, based on the thermal response characteristics under normal operating conditions, a normal thermal response feature vector is constructed. The normal thermal response feature vector includes the delay time between current disturbance and temperature rise response under normal operating conditions, the corresponding temperature rise change amplitude under normal operating conditions, the temperature rise change rate under normal operating conditions, the temperature rise duration under normal operating conditions, and the temperature recovery time under normal operating conditions.

[0106] By statistically analyzing the thermal response feature vectors corresponding to multiple normal operating samples, the variation range and feature distribution law of normal thermal response features under different load disturbance conditions are obtained, forming a normal thermal response feature library for judging abnormal heating.

[0107] S42, obtain the corresponding real-time thermal response characteristic parameters based on the current dynamic delay characteristics; Specifically, the dynamic delay characteristic parameters output in step S3 are obtained, the temperature change process corresponding to the current load disturbance event is obtained, and the real-time thermal response parameters corresponding to the normal thermal response feature library are extracted from the temperature change process. The real-time thermal response parameters include the actual delay time between the current load disturbance and the temperature rise response, the current temperature rise change amplitude, the current temperature rise change rate, the current temperature rise duration, and the current temperature recovery time. Construct the current thermal response feature vector based on real-time thermal response parameters.

[0108] S43, perform feature matching between the current thermal response characteristics and the normal thermal response characteristics to obtain a comprehensive thermal response deviation index; Specifically, feature matching is performed on the normal thermal response feature library established in step S41 and the current thermal response feature parameters obtained in step S42. Furthermore, the difference between the current thermal response feature vector and the normal thermal response feature vector is calculated to obtain the deviation of each feature parameter; Based on the importance of each thermal response characteristic parameter, the deviations are weighted and fused to obtain a comprehensive thermal response deviation index.

[0109] In the formula, This is a comprehensive thermal response deviation index of the current power metering box relative to its normal state. This represents the deviation corresponding to the i-th thermal response characteristic. Let be the weighting coefficient corresponding to the i-th thermal response feature. The weighting coefficient is determined by collecting thermal response feature data under normal operating conditions and historical abnormal heating conditions of the power metering box, calculating the correlation between the deviation of each thermal response feature parameter and the abnormality judgment result, and normalizing it according to the proportion of the correlation of each feature parameter to the sum of the correlations of all feature parameters. The number of thermal response characteristic parameters involved in the calculation of the comprehensive thermal response deviation index represents the number of thermal response characteristic dimensions used for judging abnormal heating.

[0110] S44, determine whether the power metering box has an abnormal heating fault based on the comprehensive thermal response deviation index; Specifically, the comprehensive thermal response deviation index obtained in step S43 is compared with the preset anomaly judgment threshold.

[0111] The preset anomaly judgment threshold is determined by collecting thermal response deviation index data of the power metering box under normal operating conditions and statistically analyzing the range of deviation index changes under normal conditions.

[0112] When the comprehensive thermal response deviation index does not exceed the preset abnormal judgment threshold, it means that the temperature rise response process corresponding to the current load disturbance is consistent with the normal thermal response law, and it is determined that there is no abnormal heating fault in the power metering box. When the overall thermal response deviation index is greater than the preset abnormal judgment threshold, it indicates that the current temperature rise response process deviates from the normal thermal response pattern, and the power metering box is judged to have an abnormal heating fault.

[0113] S45, based on the degree of dynamic delay anomaly and the characteristics of temperature rise change, assists in the judgment of fault type; S46, output the abnormal heating judgment result; Specifically, based on the abnormal heating state determination result obtained in step S44 and the abnormal heating fault type determination result obtained in step S45, an abnormal heating diagnosis result for the power metering box is generated. The abnormal heating diagnosis results include: abnormal heating state judgment results, which indicate whether there is an abnormal heating fault in the current power metering box; abnormal heating fault type, which indicates the fault category corresponding to abnormal heating; comprehensive thermal response deviation index, which indicates the degree of deviation of the current thermal response process from the normal thermal response law; dynamic delay abnormality degree, which indicates the degree of abnormality of the current temperature rise response delay relative to the normal state; and temperature rise characteristic abnormality degree set, which indicates the degree of abnormality of temperature rise amplitude, temperature rise rate, and temperature rise duration relative to the normal state.

[0114] Furthermore, to verify the effectiveness of abnormal fever diagnosis based on dynamic delay characteristics, a statistical analysis was performed on the degree of dynamic delay deviation between samples from normal operation and samples from abnormal fever states. The results are as follows: Figure 4 As shown.

[0115] like Figure 4As shown, under normal operating conditions, the dynamic delay deviation index is within a small range, mainly between 0 and 0.15. However, when abnormal heating faults occur, due to increased contact resistance, changes in the heat transfer path, or decreased heat dissipation capacity, the actual temperature rise response time deviates from the normal thermal response pattern, and the dynamic delay deviation index increases significantly. Therefore, by matching the dynamic delay characteristics with the normal thermal response characteristics, abnormal heating states can be effectively identified.

[0116] Furthermore, S45, based on the degree of dynamic delay anomaly and temperature rise change characteristics, assists in determining the fault type, including: S451, Calculate the degree of dynamic delay anomaly based on the difference between the current dynamic delay characteristics and the normal thermal response characteristics; The specific formula for calculating the degree of dynamic latency anomaly is as follows:

[0117] In the formula, To determine the degree of dynamic delay anomaly, This represents the actual temperature rise response delay time corresponding to the current load disturbance. This is the standard temperature rise response delay time under the same load disturbance conditions during normal operation.

[0118] Furthermore, the degree of dynamic latency anomaly is compared with a preset latency anomaly threshold; When the degree of dynamic delay anomaly exceeds the preset delay anomaly threshold, it is determined that there is a delay anomaly in the current thermal response process of the power metering box.

[0119] The preset delay anomaly threshold is determined by collecting historical dynamic delay deviation data under normal operating conditions and statistically analyzing the distribution range of dynamic delay deviation under normal conditions.

[0120] S452, calculate the degree of temperature rise anomaly based on the current temperature rise change characteristics; Specifically, based on the temperature rise amplitude, temperature rise rate, and temperature rise duration characteristics corresponding to the current temperature rise response process, these characteristics are compared with the corresponding thermal response characteristics under normal operating conditions, and the abnormality of temperature rise amplitude, abnormality of temperature rise rate, and abnormality of temperature rise duration are calculated respectively.

[0121]

[0122]

[0123] In the formula, The degree of abnormality in temperature rise amplitude, The degree of abnormality in the rate of temperature rise, The degree of abnormality in the duration of temperature rise, This represents the actual temperature rise corresponding to the current load disturbance. This represents the standard temperature rise value corresponding to normal operating conditions. Given the current rate of temperature change, This represents the standard temperature rise rate under normal operating conditions. For the current duration of temperature rise, This refers to the standard temperature rise duration under normal operating conditions.

[0124] S453, determine the type of abnormal heating based on the combination relationship between the degree of abnormal dynamic delay and the degree of abnormal temperature rise; Specifically, establish a correlation judgment rule between dynamic delay anomaly characteristics and the degree of temperature rise anomaly; When the abnormality of dynamic delay is greater than the preset abnormality threshold and the abnormality of temperature rise does not exceed the normal range threshold of temperature rise, it is determined that the current power metering box has a heat transfer abnormality or thermal inertia abnormality fault.

[0125] Abnormal heat transfer or abnormal thermal inertia indicates that the heat input generated by the load disturbance is normal, but the process of heat transfer from the heating component to the temperature detection location has changed, resulting in an increased delay in the temperature rise response time.

[0126] When the abnormal temperature rise amplitude exceeds the normal temperature rise amplitude threshold and the abnormal dynamic delay does not exceed the preset abnormal delay threshold, it is determined that there is a local abnormal heating fault in the current power metering box.

[0127] Among them, localized abnormal heating indicates that the temperature rise response time has not changed significantly, but the temperature rise amplitude generated by unit load disturbance exceeds the normal thermal response range.

[0128] When the abnormal temperature rise amplitude exceeds the normal temperature rise amplitude threshold and the abnormal temperature rise rate exceeds the preset abnormal temperature rise rate threshold, it is determined that the current power metering box has a rapid heat accumulation abnormal fault.

[0129] Rapid heat accumulation anomaly indicates that there is an increase in contact resistance inside the equipment, increased heating of connecting parts, or decreased heat dissipation capacity, leading to a rapid rise in temperature.

[0130] When the abnormality of dynamic delay exceeds the preset abnormality threshold, and the abnormality of temperature rise duration exceeds the preset abnormality threshold, it is determined that the current power metering box has an abnormal fault of reduced heat dissipation capacity.

[0131] An abnormal decrease in heat dissipation capacity indicates that the equipment releases heat at a slower rate, causing the temperature rise to persist for a longer period than normal.

[0132] Among them, the normal range threshold for temperature rise amplitude is determined by collecting historical temperature rise amplitude data under normal operating conditions of the power metering box and statistically analyzing its maximum allowable deviation range; the preset abnormal temperature rise rate threshold is determined by statistically analyzing the maximum variation range of historical temperature rise change rate under normal operating conditions; and the abnormal temperature rise duration threshold is determined by statistically analyzing the distribution range of historical temperature rise duration under normal operating conditions and selecting its maximum allowable duration.

[0133] S454, When multiple abnormal conditions are met simultaneously, the main fault type is determined according to the principle of the greatest degree of abnormality; Specifically, the comprehensive anomaly score corresponding to each fault type is calculated separately:

[0134] In the formula, The anomaly score corresponding to the k-th type of abnormal fault. , , , These are the weight coefficients for the corresponding abnormal features.

[0135] The abnormal scores corresponding to each fault type are compared, and the fault type with the highest score is determined as the current main abnormal heating type.

[0136] The weighting coefficient is obtained by collecting historical abnormal heating fault samples, acquiring the change amount corresponding to each abnormal feature parameter, and calculating the number of matches between the change amount of each abnormal feature parameter and the corresponding fault type. The proportion of the number of matches of each abnormal feature parameter to the total number of matches of all abnormal feature parameters is used as the corresponding weighting coefficient.

[0137] It should be noted that by establishing a thermal response feature library under normal operating conditions of the power metering box, and by performing multi-dimensional matching of the dynamic delay features and temperature rise change features corresponding to the current load disturbance with the normal thermal response law, the dynamic correlation between current disturbance intensity, heat accumulation process, equipment thermal inertia, and temperature rise response features can be comprehensively considered. Compared with methods that rely solely on temperature thresholds or single delay times for anomaly judgment, this improves the accuracy and adaptability of abnormal heating detection. Simultaneously, by constructing a comprehensive thermal response deviation index and weighting and fusing dynamic delay deviation, temperature rise amplitude deviation, temperature rise rate deviation, and temperature rise duration deviation, the impact of ambient temperature changes, load fluctuations, and equipment operating status changes on the detection results can be reduced, improving the reliability of abnormal heating fault identification. Furthermore, by combining the abnormality degree of dynamic delay with the abnormality degree of temperature rise change, different fault types such as abnormal heat transfer or thermal inertia, localized abnormal heating, rapid heat accumulation, and decreased heat dissipation capacity can be differentiated. This allows the detection results to not only determine the abnormal heating state but also provide a basis for anomaly cause analysis, thereby improving the accuracy, interpretability, and engineering application value of power metering box fault diagnosis.

[0138] Reference Figure 2 As shown in the schematic diagram, the IoT-based power metering box fault diagnosis system provided by this invention includes: a load disturbance identification module, a temperature rise response analysis module, a dynamic delay feature construction module, and an abnormal heating diagnosis module. The modules are interconnected. The load disturbance identification module is used to identify load disturbance events based on the current change characteristics during the operation of the power metering box, and to determine the time when the load disturbance occurs. The temperature rise response analysis module is used to analyze the temperature response process within a preset time range based on the time of load disturbance, and to determine the time of temperature rise response caused by load disturbance. The dynamic delay feature construction module is used to construct the dynamic delay feature between current disturbance and temperature rise response based on the time difference between the occurrence of load disturbance and the occurrence of temperature rise response. The abnormal heating diagnosis module is used to match the dynamic delay characteristics with the corresponding thermal response characteristics under normal operating conditions of the power meter box, and determine whether there is an abnormal heating fault in the power meter box based on the matching results.

[0139] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0140] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0141] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0142] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A fault diagnosis method for electricity metering boxes based on the Internet of Things, characterized in that, include: Identify load disturbance events based on the current change characteristics during the operation of the power metering box, and determine the time when the load disturbance occurs; Based on the moment of load disturbance, the temperature response process within a preset time range is analyzed to determine the moment of temperature rise response caused by load disturbance; Based on the time difference between the occurrence of load disturbance and the temperature rise response, a dynamic delay characteristic between current disturbance and temperature rise response is constructed. The dynamic delay characteristics are matched with the thermal response characteristics corresponding to the normal operation of the power meter box, and the matching results are used to determine whether the power meter box has an abnormal heating fault.

2. The method according to claim 1, characterized in that, The process of identifying load disturbance events based on current change characteristics during the operation of the power metering box and determining the time of occurrence of the load disturbance includes: Collect continuous current monitoring data, construct a current change time series dataset according to the sampling time sequence, and extract the current disturbance change characteristics; The current disturbance change characteristics are matched with preset disturbance judgment conditions. When the preset disturbance judgment conditions are met, it is identified as a load disturbance event, and the starting sampling time of the continuous current change interval that meets the conditions is taken as the candidate occurrence time of the load disturbance. Time correction is performed on the candidate occurrence times of load disturbances to determine the actual occurrence time of the load disturbance.

3. The method according to claim 2, characterized in that, The step of time-correcting the candidate occurrence times of load disturbances to determine the occurrence time of load disturbances includes: When the candidate occurrence time of load disturbance does not meet the preset disturbance judgment conditions, the search proceeds sequentially along the current sampling time sequence, and the sampling time that first simultaneously meets all the judgment conditions of current change amplitude, change rate and duration is determined as the corrected disturbance occurrence time. Using the corrected disturbance occurrence time as a benchmark, the consistency of the current change data within the subsequent preset time window is verified. If the verification passes, the time is output as the final disturbance occurrence time; if the verification fails, the search continues along the time sequence to find the next sampling time that meets all the judgment conditions until the consistency verification is passed.

4. The method according to claim 1, characterized in that, The determination of the temperature rise response time caused by the load disturbance includes: Obtain the temperature rise response time range and determine the candidate time of the temperature rise response; The time interval between the candidate time of temperature rise response and the time of load disturbance is matched with the preset thermal response time range. At the same time, it is determined whether there is a positive correlation between the current change amplitude of the load disturbance and the temperature rise change. If so, it is confirmed that the temperature rise change is caused by the load disturbance. Based on the confirmed candidate time of temperature rise response, search along the temperature sampling time sequence for the first sampling time that simultaneously meets all the judgment conditions of temperature rise change, temperature rise rate and temperature rise duration, and correct it as the time of temperature rise response. The consistency of temperature change data within a preset range after the corrected temperature rise response time is verified. If the verification is successful, the time is output as the final temperature rise response time; otherwise, the search continues along the time sequence until the verification is successful.

5. The method according to claim 4, characterized in that, The range of time for obtaining the temperature rise response includes: Calculate the disturbance heat characterization parameters based on the current change data before and after the load disturbance occurs; The initial temperature rise response time range is obtained by matching historical load disturbance samples with disturbance heat characterization parameters. Based on temperature change data within the initial temperature rise response time range, the thermal inertia parameters of the power metering box are identified; The time range of heat influence is determined based on the heat disturbance characterization parameters, and the time range of thermal inertia influence is determined based on the thermal inertia parameters. The time range of heat influence and the time range of thermal inertia influence are then merged to obtain the dynamic temperature rise response time range.

6. The method according to claim 1, characterized in that, The dynamic delay characteristics between the constructed current disturbance and the temperature rise response include: The base delay time is obtained based on the time difference between the occurrence of the load disturbance and the temperature rise response time. Based on the dynamic delay characteristic data of historical load disturbance events, a dynamic delay mapping relationship between current disturbance characteristics and temperature rise response delay time is established. Input the current disturbance characteristics corresponding to the current load disturbance into the dynamic delay mapping relationship to obtain the theoretical delay time corresponding to the current load disturbance; The basic delay time is compared with the theoretical delay time. When the deviation between the two is within the preset allowable range, the current dynamic delay feature is confirmed to be valid. The confirmed valid dynamic delay feature is then used to identify the abnormal heating state of the power metering box.

7. The method according to claim 6, characterized in that, The establishment of a dynamic delay mapping relationship between current disturbance characteristics and temperature rise response delay time includes: An initial mapping model between current disturbance characteristic parameters and temperature rise response delay time is established based on the historical dynamic delay characteristic set. Obtain the deviation between the current thermal state parameters of the equipment and the standard thermal state parameters under normal operating conditions, and determine the thermal state correction coefficient based on the deviation. The initial mapping model is corrected using thermal state correction coefficients to obtain the dynamic delay mapping relationship after adaptive correction of the equipment thermal state.

8. The method according to claim 1, characterized in that, The step of matching the dynamic delay characteristics with the thermal response characteristics corresponding to the normal operating state of the power metering box, and determining whether the power metering box has an abnormal heating fault based on the matching result, includes: Obtain the real-time thermal response characteristic parameters corresponding to the current load disturbance, and compare them with the normal thermal response characteristics to obtain the deviation of each thermal response characteristic parameter; The various deviations are weighted and fused to obtain a comprehensive thermal response deviation index; When the deviation of the comprehensive thermal response index does not exceed the preset abnormal judgment threshold, it is determined that there is no abnormal heating fault; otherwise, it is determined that there is an abnormal heating risk, and the fault type is assisted in judging based on the degree of abnormality of dynamic delay characteristics and temperature rise change characteristics, and the abnormal heating judgment result is output.

9. The method according to claim 8, characterized in that, The method of assisting in fault type determination based on the degree of abnormality of dynamic delay characteristics and temperature rise change characteristics includes: The deviation between the actual temperature rise response delay time corresponding to the current load disturbance and the standard temperature rise response delay time under normal operating conditions is used as the degree of dynamic delay anomaly. The deviation between the temperature rise change characteristics during the current temperature rise response and the corresponding standard value under normal operating conditions is obtained as the degree of temperature rise anomaly. The type of abnormal heating is determined based on the combined relationship between the degree of dynamic delay anomaly and the degree of temperature rise anomaly. When the dynamic delay is abnormal but the temperature rise is normal, it is determined to be an abnormal heat transfer or an abnormal thermal inertia. When the temperature rise amplitude is abnormal but the dynamic delay is normal, it is determined to be local abnormal heating; When the temperature rise amplitude and temperature rise rate are abnormal, it is determined to be an abnormal rapid heat accumulation. When the dynamic delay and the duration of temperature rise are abnormal, it is determined that the heat dissipation capacity is abnormally reduced.

10. A system using the Internet of Things-based fault diagnosis method for power meter boxes as described in any one of claims 1-9, characterized in that, include: The load disturbance identification module is used to identify load disturbance events based on the current change characteristics during the operation of the power metering box, and to determine the time when the load disturbance occurs. The temperature rise response analysis module is used to analyze the temperature response process within a preset time range based on the time of load disturbance, and to determine the time of temperature rise response caused by load disturbance. The dynamic delay feature construction module is used to construct the dynamic delay feature between current disturbance and temperature rise response based on the time difference between the occurrence of load disturbance and the occurrence of temperature rise response. The abnormal heating diagnosis module is used to match the dynamic delay characteristics with the thermal response characteristics corresponding to the normal operation of the power meter box, and to determine whether there is an abnormal heating fault in the power meter box based on the matching results.

Images