Cable terminal fault diagnosis method and system based on infrared online monitoring

Abstract

This invention discloses a method and system for diagnosing cable terminal faults based on infrared online monitoring, belonging to the field of terminal fault diagnosis technology. The method involves acquiring visible light and infrared thermal images of the cable terminal area and overlaying them to generate a dual-light fusion image. Based on pre-calibrated components of the cable terminal area, at least one monitoring region is divided on the dual-light fusion image to obtain a set of monitoring regions and component labels for each component of the cable terminal. Temperature characteristic parameters of the monitoring regions are extracted and dynamically filtered to obtain temperature trend values. Deviation comparisons are performed on the obtained temperature trend values ​​to generate a comprehensive deviation marker set for determining the validity of the terminal fault. This method effectively solves the problem of low accuracy in identifying faulty cables when the cable terminal is operating under low load in existing technologies.

Description

Technical Field

[0001] This invention relates to the field of terminal fault diagnosis technology, and in particular to a method and system for diagnosing cable terminal faults based on infrared online monitoring. Background Technology

[0002] Infrared online monitoring is a very practical technology in power condition maintenance. It utilizes the abnormal temperature changes in cable terminals, such as outdoor terminals and GIS terminals, caused by dielectric loss and increased contact resistance before a fault occurs. This abnormal temperature change is monitored non-contactly using an infrared thermal imager, and combined with a series of analytical algorithms to determine the equipment status. Its specific implementation process typically includes the following steps: Infrared thermal imagers, fixedly installed within the substation, are used to monitor cable terminals, simultaneously recording ambient temperature, humidity, wind speed, and load current. To ensure the accuracy of temperature data, clear infrared images and a fixed shooting angle are essential. The acquired raw thermal images undergo image preprocessing, including noise reduction, correction, and alignment with a reference image using image registration techniques. This identifies and delineates key parts of the cable terminal, including but not limited to conductor terminals reflecting connection conditions, stress cone locations indicating internal insulation or stress cone defects, and grounding connection points. Temperature characteristic parameters of each part of the cable terminal are automatically extracted, and the temperature rise in each area is calculated to determine the presence of overheating defects. A horizontal comparison is made between the temperatures of corresponding locations of three-phase cable terminals in the same circuit, and a longitudinal trend analysis is performed, comparing the current temperature with the cable terminal's baseline. For current-induced heating defects under low load conditions, a relative temperature difference method is applied, assessing the degree of heating by calculating the percentage temperature difference between defective and non-defective points. The comprehensive analysis results provide different levels of fault assessment, including critical defects.

[0003] However, infrared online monitoring essentially calculates temperature by receiving infrared energy radiated from the surface of an object. This is greatly affected by the on-site environment. For example, direct sunlight will cause the surface of the cable terminal to heat up additionally, while strong winds or rain and snow will accelerate surface heat dissipation, creating a forced cooling effect. During light load periods, the current is usually less than the rated value. The heat generated by the cable terminal is closely related to the square of the load current (for resistive losses) or the voltage (for dielectric losses). Therefore, in actual power grid operation, the absolute temperature rise generated at this time is also relatively small. Consequently, in light load current application scenarios, the signal strength for some early temperature rises is often weakened by the load reduction, making it difficult to effectively identify faulty cables due to the small absolute temperature rise amplitude. This results in a problem of low accuracy in identifying faulty cables during low load operation. Summary of the Invention

[0004] To address the issue of low accuracy in identifying faulty cables during low-load operation in existing technologies, this invention provides a cable terminal fault diagnosis method and system based on infrared online monitoring. The technical solution is as follows: On the one hand, a cable terminal fault diagnosis method based on infrared online monitoring is provided. This method includes: S1, acquiring visible light and infrared thermal images of the cable terminal area and overlaying them to generate a dual-light fusion image; S2, based on pre-calibrated components of the cable terminal area, dividing at least one monitoring area on the dual-light fusion image to obtain a set of monitoring areas and component labels for each component of the cable terminal; the monitoring areas include at least one or more of the following: conductor connection or crimping point area, insulation outer surface area, and terminal head or sleeve outer surface area; S3, extracting temperature characteristic parameters of the monitoring area and performing dynamic trend filtering on the temperature characteristic parameters to obtain temperature trend values; the temperature characteristic parameters include at least one or more of the following: highest temperature of the monitoring area, average temperature of the monitoring area, heating rate, and direction; the temperature trend value is used to characterize the stable change direction and magnitude of temperature over time within the monitoring area; S4, performing deviation comparison judgment on the obtained temperature trend values ​​to generate a comprehensive deviation marker set for terminal fault validity determination; the comprehensive deviation marker set includes at least one or more of the following: absolute over-temperature deviation marker, relative temperature rise deviation marker, abnormal heating rate deviation marker, and abnormal temperature gradient deviation marker.

[0005] On the other hand, a cable terminal fault diagnosis system based on infrared online monitoring is provided. This system includes: a dual-light fusion module, used to acquire visible light images and infrared thermal images of the cable terminal area and overlay the images to generate a dual-light fusion image; a monitoring area division module, used to divide at least one monitoring area on the dual-light fusion image based on pre-calibrated components of the cable terminal area, to obtain a set of monitoring areas and component labels for each component of the cable terminal; a dynamic trend filtering module, used to extract temperature feature parameters of the monitoring area and perform dynamic trend filtering on the temperature feature parameters to obtain temperature trend values; and a deviation comparison and judgment module, used to perform deviation comparison and judgment on the obtained temperature trend values, thereby generating a comprehensive deviation label set for determining the validity of the terminal fault.

[0006] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. This invention achieves high-precision spatial alignment of visible light and infrared thermal images by simultaneously acquiring visible light and infrared images at the same acquisition time and introducing image registration technology based on scale-invariant feature transformation and bidirectional matching strategy. Unlike traditional simple overlay or manual registration methods, this scheme calculates the feature point sets of visible light and infrared images before image overlay. Reliable matching point pairs are selected through descriptor similarity calculation and reverse optimal matching verification. A perspective transformation model is then established, and the spatial transformation matrix is ​​solved using the least squares method, enabling the infrared image to be accurately mapped to the visible light image coordinate system. After interpolation and resampling, the transformed infrared image is the same size and spatially aligned with the original visible light image. The resulting dual-light fused image retains the texture details of the visible light image while embedding infrared temperature information. Based on this fused image, the monitoring areas of each component are automatically divided according to a pre-calibrated template image, generating a component coordinate mapping table. This enables precise positioning and intelligent identification of key parts such as conductor connection points, insulating surfaces, and terminal heads, providing an accurate spatial reference for subsequent temperature feature extraction and fault determination, and improving the reliability of monitoring data.

[0007] 2. Based on the division of monitoring areas, this invention introduces a hierarchical monitoring mechanism. By calculating regional fluctuation indicators, the monitoring area is divided into a first deviation level and a second deviation level. For critical areas with large fluctuations, high-frequency sampling and Kalman filtering are used to generate a high-time-density temperature trend value sequence. For non-critical areas with smaller fluctuations, low-frequency sampling and exponentially weighted moving average are used to generate stable temperature trend values. This differentiated processing method not only effectively reduces the overall computational burden but also improves the data acquisition density and real-time performance of trend analysis in critical areas. It can not only track the inherent laws of temperature changes in real time but is also suitable for smoothing stable areas. Compared with the traditional unified sampling and fixed filtering strategy, this invention can capture abnormal temperature changes in potential fault areas earlier, avoiding missed or delayed alarms caused by insufficient sampling frequency, and achieving efficient allocation of monitoring resources and accurate triggering of fault warnings.

[0008] 3. After acquiring the temperature trend value of the monitoring area, this invention constructs a multi-level deviation comparison and judgment mechanism. Through three-phase deviation comparison, the absolute difference between each pair of temperature trend values ​​at the same location in the three phases at the same time is calculated, and the maximum difference is taken as the maximum temperature difference between the three phases. Combined with a preset reference temperature difference, a three-phase deviation coefficient is generated. This step emphasizes lateral comparison between the three phases, enabling the identification of temperature anomalies caused by poor contact, uneven load, or local faults. This avoids misjudgments that may arise from relying solely on absolute temperature or single-phase historical data. Furthermore, adaptive deviation judgment is introduced for each monitoring area. Based on the statistical mean and standard deviation of the historical temperature trend value sequence, a dynamic adaptive reference interval is constructed using a preset confidence coefficient. This interval can automatically adjust with factors such as load fluctuations, avoiding failure during long-term operation. The real-time temperature trend value is compared with the reference interval to identify high or low adaptive exceedances, and the deviation magnitude is recorded. This dual judgment mechanism, combining lateral comparison and vertical adaptation, considers both the consistency of similar operations and the characteristics of individual operations, improving the accuracy of fault judgment and making it suitable for the effective capture of early, minor anomalies. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A flowchart of a cable terminal fault diagnosis method based on infrared online monitoring provided in an embodiment of the present invention; Figure 2 A flowchart corresponding to image overlay provided in an embodiment of the present invention; Figure 3 A flowchart corresponding to the division of monitoring areas provided in this embodiment of the invention; Figure 4 The flowchart corresponding to the deviation comparison judgment provided in the embodiments of the present invention; Figure 5 This is a schematic diagram of the structure of a cable terminal fault diagnosis system based on infrared online monitoring provided in an embodiment of the present invention. Detailed Implementation

[0011] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0012] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0013] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0014] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0015] This invention provides a cable terminal fault diagnosis method based on infrared online monitoring, such as... Figure 1 The flowchart shown is for a cable terminal fault diagnosis method based on infrared online monitoring. The processing flow of this method may include the following steps: S1, acquiring visible light images and infrared thermal images of the cable terminal area and overlaying the images to generate a dual-light fusion image; S2, based on pre-calibrated components of the cable terminal area, dividing at least one monitoring area on the dual-light fusion image to obtain a set of monitoring areas and component labels for each component of the cable terminal; the monitoring areas include at least one or more of the following: conductor connection or crimping point area, insulation outer surface area, terminal head or sleeve outer surface area; S3, extracting temperature characteristic parameters of the monitoring area and performing dynamic trend filtering on the temperature characteristic parameters to obtain temperature trend values; the temperature characteristic parameters include at least one or more of the following: highest temperature of the monitoring area, average temperature of the monitoring area, heating rate and direction, and the temperature trend value is used to characterize the stable change direction and magnitude of temperature over time in the monitoring area; S4, performing deviation comparison judgment on the obtained temperature trend values ​​to generate a comprehensive deviation mark set for terminal fault validity determination; the comprehensive deviation mark set includes at least one or more of the following: absolute over-temperature deviation mark, relative temperature rise deviation mark, abnormal heating rate deviation mark, and abnormal temperature gradient deviation mark.

[0016] like Figure 2 The diagram shows a flowchart of image overlay provided in an embodiment of the present invention. In step S1, image overlay is used to spatially align and fuse a visible light image and an infrared thermal image acquired at the same time to generate a dual-light fused image containing texture details and temperature information. The specific process includes the following steps: At the same acquisition time, visible light imaging and infrared thermal imaging are triggered to work synchronously to acquire the original visible light image and the original infrared image of the cable terminal area. To ensure the accuracy of subsequent registration, the acquisition of the two images must be synchronized in time to avoid image misalignment caused by cable terminal jitter or environmental changes.

[0017] The scale-invariant feature transform (SIFT) operation is performed on the original visible light image to extract significant feature points with scale invariance and rotation invariance, forming a set of visible light feature points. These feature points typically correspond to parts with obvious texture features, such as the outline edge of the cable terminal, the corner point of the hardware, and the edge of the insulator skirt.

[0018] The same SIFT operation is performed on the original infrared image to extract feature points and form an infrared feature point set. Since the infrared image has low resolution and little texture information, the parameters of the SIFT algorithm need to be adaptively adjusted. The adjustment value is determined based on the average gradient ratio between the infrared image and the visible light image. For example, the contrast threshold is reduced from the default 0.04 to 0.02-0.03 to ensure that a sufficient number of reliable feature points are extracted.

[0019] Based on feature descriptors, bidirectional matching of the visible light feature point set and the infrared feature point set is performed, including verification in the following two directions: Forward matching: For each visible light feature point, calculate the Euclidean distance between its feature descriptor and the feature descriptors of each point in the infrared feature point set, and record the infrared feature point that is closest to the visible light feature point descriptor as a candidate matching point.

[0020] Reverse matching: For a candidate matching point, calculate the Euclidean distance between its feature descriptor and the feature descriptors of each point in the set of visible light feature points. If the visible light feature point that is closest to the candidate matching point's descriptor is exactly the corresponding visible light feature point, it means that the matching relationship satisfies bidirectional consistency; otherwise, it is considered a mismatch and is removed.

[0021] For matching point pairs that satisfy bidirectional consistency, their descriptor similarity is further calculated, quantized using the inverse of the distance value, and matching point pairs with similarity greater than a preset similarity threshold are retained to form an initial set of matching points. The preset similarity threshold is represented by the result of summing and averaging historical similarities.

[0022] The coordinates (x) of the visible light feature points in the initial matching point set k v y k v ) and corresponding infrared feature point coordinates (x ki y k i Using these as constraint samples, a perspective transformation model is established between the visible light image coordinate system and the infrared image coordinate system: ; Where H is the spatial transformation matrix to be solved, h 33 =1 is used for normalization; the least squares method is used to solve the parameters of the spatial transformation matrix to minimize the sum of squared mapping errors of all matching point pairs, that is, to solve the optimization problem. The above nonlinear least squares problem is solved by numerical optimization methods, such as the Levenberg-Marquardt algorithm, to obtain the matrix element values ​​of the spatial transformation matrix H.

[0023] Based on the calculated spatial transformation matrix H, the coordinates of each pixel in the original infrared image are transformed, and its corresponding position in the visible light image coordinate system is calculated. Since the transformed coordinates may be non-integer positions, the bilinear interpolation method is used to resample the transformed image to generate a transformed infrared image with the same size and spatial alignment as the original visible light image.

[0024] The transformed infrared image is overlaid onto the corresponding area of ​​the original visible light image in the form of a semi-transparent pseudo-color overlay to generate a dual-light fusion image. The overlay method can use a fixed transparency, such as setting the infrared layer transparency to 30%, or an adaptive transparency to ensure that visible light texture details and infrared temperature information are clearly presented at the same time.

[0025] like Figure 3 The diagram shows a flowchart corresponding to the monitoring area division provided in an embodiment of the present invention. In step S2, the monitoring area is divided based on the generated dual-light fusion image, which specifically includes the following sub-steps: S201, Acquire template image and component calibration: Acquire a clear dual-light fusion image as a template image; On this template image, calibrate the outline range of each key component of the cable terminal through manual interaction or automatic recognition; Key components include at least the conductor connection or crimping point area, the insulation outer surface area, and the terminal head or sleeve outer surface area; Each calibrated component outline is represented by a set of polygon vertex coordinates and is assigned a unique component label, which contains information such as component type and phase.

[0026] S202, Generate a component coordinate mapping table: Save the contour coordinates of each component calibrated in step S201 to the local database to form a component coordinate mapping table. This mapping table contains the following fields: component ID, component name, phase, contour vertex coordinate sequence, and calibration timestamp. Once established, the mapping table can be reused in all subsequent inspection cycles to ensure that the area divided during each monitoring session remains consistent with the template image.

[0027] S203, Initial monitoring area division: After obtaining a new dual-light fusion image in each inspection cycle, read the coordinates of the contour vertices of each component stored in the component coordinate mapping table, draw the corresponding monitoring area in polygon form on the dual-light fusion image, and generate an initial monitoring area set that corresponds one-to-one with the component label. Each monitoring area contains the following attributes: area ID, component label, pixel coordinate range, and current temperature feature parameter storage area.

[0028] S204, Calculate the regional fluctuation index: For each initial monitoring area, calculate the following two basic indicators: Fault Frequency Coefficient: This coefficient is obtained by normalizing the statistical value of the number of temperature anomaly alarms that occurred in the corresponding component of the region during the historical operating cycle. The normalization method is as follows: the maximum and minimum values ​​of historical alarm counts in all regions are statistically analyzed, and the difference between the historical alarm count and the minimum value in the region is used as the numerator, while the difference between the maximum and the minimum value is used as the denominator. The coefficient ranges from 0 to 1, with a larger value indicating that the region is more prone to faults historically.

[0029] Temperature fluctuation coefficient: It is determined based on the ratio of the temperature standard deviation of the area during the inspection cycle to the preset fluctuation benchmark value. The coefficient ranges from 0 to 1. The larger the value, the more severe the temperature fluctuation in the area. The preset fluctuation benchmark value is the result of summing and averaging the historical temperature standard deviations.

[0030] The regional fluctuation index is obtained by multiplying the fault frequency coefficient and temperature fluctuation coefficient by the corresponding weight coefficients and summing them. The preset weight coefficients satisfy the condition that the sum of the weight coefficients is 1, and the default configuration is the preset weight coefficient = 0.5. The preset personnel can change it according to the current working conditions.

[0031] If the regional fluctuation index is greater than the preset regional fluctuation index, the corresponding initial monitoring area is recorded as the first deviation level area. Temperature feature parameters are repeatedly extracted during the inspection cycle, and the extracted temperature feature parameter sequence is smoothed by Kalman filtering to obtain a high time density first temperature trend value sequence.

[0032] Conversely, the corresponding initial monitoring area is designated as the second deviation level area, and temperature characteristic parameters are extracted once within the inspection cycle. The extracted temperature characteristic parameters are then smoothed using an exponentially weighted moving average to obtain a stable second temperature trend value. The values ​​of the first and second sampling frequencies are preset according to the deviation level of the monitoring area. In this embodiment, the first sampling frequency is set to once every 15 minutes, and the second sampling frequency is set to once every 2 hours. These specific values ​​can be adjusted according to actual monitoring needs and operational experience, as long as the first sampling frequency is greater than the second sampling frequency. After the sampling frequency is determined, the temperature characteristic parameter extraction operation is performed according to this frequency in subsequent inspection cycles.

[0033] For the first deviation level region, the temperature trend value sequence is obtained using the first sampling frequency. When making deviation comparison judgments, in addition to using the latest temperature trend value, dynamic characteristics such as heating rate and acceleration can also be calculated based on the sequence data, and trend prediction can be performed. The heating rate anomaly judgment is only performed for the first deviation level region and is calculated based on the temperature difference between adjacent sampling points.

[0034] For the second deviation level region, the temperature trend value is obtained using the second sampling frequency, and only the latest temperature trend value is used for absolute over-temperature deviation judgment and adaptive deviation judgment.

[0035] Through the aforementioned hierarchical monitoring mechanism, differentiated configuration of monitoring resources for different components of the cable terminal is achieved. For critical areas with historically frequent faults or severe temperature fluctuations, high-frequency sampling and Kalman filtering are used for fine tracking, which can capture abnormal temperature rise trends earlier. For stable areas, low-frequency sampling is used to reduce the computational burden. This adaptive allocation strategy effectively reduces overall power consumption and data processing pressure while ensuring monitoring sensitivity.

[0036] like Figure 4 The diagram shows a flowchart of the deviation comparison judgment provided in an embodiment of the present invention. In actual cable lines, cables in the same circuit typically contain three phases, denoted as phase A, phase B, and phase C. The cable terminals of each phase, including outdoor terminals, GIS terminals, etc., are physically independent but together constitute a complete power transmission circuit. In step S2, the same monitoring area, such as the conductor connection area, has been divided for each phase cable terminal, and each monitoring area has been assigned a component label containing phase information.

[0037] Within the current inspection cycle, based on the phase information in the component label, extract the current temperature trend values ​​of phase A, phase B, and phase C in the same monitoring area from the dual-light fusion image at the same sampling time. The three temperature trend values ​​must come from the same sampling time, the same type of monitoring area (e.g., all are conductor connection areas), and the division method of the monitoring area must be consistent.

[0038] Based on the obtained three-phase temperature trend values, the absolute differences between each pair of the three phases are calculated. The maximum value among the three absolute differences is taken as the maximum interphase temperature difference, which characterizes the current degree of three-phase temperature imbalance. To quantify the severity of the temperature imbalance, this method introduces an interphase deviation coefficient. This coefficient is determined based on the ratio between the maximum interphase temperature difference and a preset reference temperature difference value. The preset reference temperature difference value can be set according to the voltage level, material characteristics, and operating experience of the cable terminal. For example, for a 110kV cable terminal, the reference temperature difference value can be set to 5℃. The value range of the interphase deviation coefficient is 0 to 1. The larger the value, the more severe the three-phase temperature imbalance.

[0039] If the maximum interphase temperature difference exceeds the preset interphase temperature difference threshold, the current inspection cycle is deemed to have triggered interphase deviation. In this case, based on the relationship between the temperature trend values ​​of the two phases contributing to the maximum interphase temperature difference, the phase with the higher temperature trend value is identified as the abnormal phase. An interphase deviation marker is generated for this abnormal phase, and the deviation magnitude is recorded. Conversely, if the temperature trend value is lower, the current inspection cycle is deemed not to have triggered interphase deviation, and no interphase deviation marker is generated. The preset interphase temperature difference threshold is represented by collecting historical temperature data from the cable terminals, statistically analyzing the temperature difference distribution in the monitoring areas of the same location across the three phases at the same time, and calculating the average of all temperature difference values. This threshold can be manually adjusted based on operational experience.

[0040] By using the above-mentioned three-phase deviation judgment, a lateral temperature comparison of the same location at the three-phase terminals of the same circuit cable is achieved. Since the three-phase terminals are under the same environmental conditions, the influence of environmental factors on the three-phase temperature is basically the same. The calculation of the three-phase temperature difference can effectively offset environmental interference and highlight the temperature anomalies caused by internal faults.

[0041] In step S4, the deviation comparison judgment also includes adaptive deviation judgment, which is used to construct a dynamic reference benchmark through the historical operating data of a single monitoring area to identify abnormal temperature changes. The specific implementation process is as follows: For each monitoring area, the system maintains its historical temperature trend value sequence in a local or cloud database. When performing adaptive deviation judgment, the system first retrieves the historical temperature trend value sequence of the monitoring area within a preset historical window from the historical database. The preset historical window is used to limit the range of historical data participating in the statistical analysis and can be set according to the operating characteristics of the cable terminal and monitoring requirements.

[0042] In this embodiment, the preset historical window uses the data window of the most recent 30 days, and abnormal data points caused by alarms or maintenance are removed to ensure that the historical data included in the statistics can represent the normal operating status.

[0043] Statistical analysis is performed on the acquired historical temperature trend value series to calculate its statistical mean and statistical standard deviation. To further optimize the adaptability of the reference interval, this method introduces an adaptive deviation coefficient, which reflects the relationship between the temperature fluctuation characteristics of the current monitoring area and the overall historical fluctuation characteristics. It is obtained by processing the ratio of the statistical standard deviation to the historical temperature standard deviation. The historical temperature standard deviation is the overall statistical measure of temperature fluctuation in the monitoring area over a longer historical period. The coefficient ranges from 0 to 1. The larger the value, the more severe the temperature fluctuation within the current window, and the reference interval should be widened accordingly to avoid too many false alarms during periods of large fluctuations.

[0044] An adaptive reference interval for the monitoring area in the current inspection cycle is constructed based on the statistical mean, statistical standard deviation, adaptive deviation coefficient, and preset confidence coefficient. Its upper limit is defined as the sum of the products of the statistical mean, statistical standard deviation, adaptive deviation coefficient, and preset confidence coefficient. Its lower limit is defined as the difference between these three products. The preset confidence coefficient determines the width of the reference interval and can be set according to the trade-off between monitoring sensitivity and false alarm rate. Under the assumption of a normal distribution, a preset confidence coefficient of 2 corresponds to approximately a 95% confidence level, and a preset confidence coefficient of 3 corresponds to approximately a 99.7% confidence level. In this embodiment, a stricter confidence level is adopted for the first deviation level area of ​​key concern, with a preset confidence coefficient of 2.5; for the second deviation level area of ​​routine concern, a preset confidence coefficient of 3 is set. This coefficient can also be manually adjusted based on operational experience.

[0045] Obtain the current temperature trend value of the monitoring area within the current inspection cycle and compare it with the constructed adaptive reference interval: if the temperature trend value is greater than the upper limit of the adaptive reference interval, adaptive deviation is triggered, and the deviation type is recorded as high adaptive overlimit and its corresponding deviation magnitude; if the temperature trend value is less than the lower limit of the adaptive reference interval, adaptive deviation is triggered, and the deviation type is recorded as low adaptive overlimit and its corresponding deviation magnitude; if the temperature trend value is within the adaptive reference interval, adaptive deviation is not triggered.

[0046] Through the above adaptive deviation judgment, dynamic monitoring of a single monitoring area based on its own historical operating patterns is realized. The reference interval is dynamically updated with historical data, which can automatically adapt to changes in long-term operating conditions such as load changes and seasonal changes, and avoid the failure of fixed thresholds in long-term operation. Moreover, each monitoring area constructs an independent reference interval based on its own historical data, taking into account the individual differences of different components and different phases, making the judgment more targeted.

[0047] In step S4, a deviation comparison judgment is performed based on the temperature trend value obtained in step S3 to generate a comprehensive deviation marker set, and the validity of the terminal fault is determined based on this marker set. The comprehensive deviation marker set is a summary of multiple deviation judgment results and includes at least one or more of the following deviation markers: Absolute over-temperature deviation marker, i.e., to judge absolute over-temperature deviation: generated based on whether the temperature trend value of the monitored area exceeds the preset absolute temperature threshold, used to identify critical situations where the temperature has exceeded the upper limit of safe operation of the equipment. The preset absolute temperature threshold is set according to the maximum allowable operating temperature of the cable terminal. For example, for cross-linked polyethylene cable terminals, it can be set to 90℃ according to relevant standards such as GB / T11017.3; or it can be determined according to the technical parameters provided by the equipment manufacturer.

[0048] Relative temperature rise deviation marker: generated based on whether the maximum temperature difference between the three phases exceeds the preset relative temperature difference threshold in the three-phase deviation judgment. It is used to identify the relative temperature rise abnormality caused by the three-phase imbalance. The preset relative temperature difference threshold is set according to the requirements of the relative temperature difference criterion in DL / T664.

[0049] An abnormal deviation in the heating rate is flagged, which is used to determine if the heating rate is abnormal. This is based on whether the heating rate of the monitored area exceeds a preset rate threshold. It is used to identify early signs of a rapid temperature rise. The preset rate threshold is determined based on the statistical distribution of the heating rate under historical normal operating conditions. The threshold is taken as three times the standard deviation of the historical heating rate average, or it can be set to 5℃ / h based on operating experience.

[0050] Temperature gradient anomaly deviation marker: This marker is generated based on whether the area ratio of hot spots or the temperature dispersion in the monitoring area exceeds a preset gradient threshold. It is used to identify abnormal states such as local overheating or uneven temperature distribution. The preset gradient threshold is determined based on the difference between the highest temperature and the average temperature in the monitoring area or the area ratio of hot spots. It can be set according to the cable terminal type. For example, for conductor connection areas, the temperature gradient threshold can be set to 10℃.

[0051] Within the current inspection cycle, based on the obtained deviation judgment results, a comprehensive deviation confidence score is calculated to quantify the credibility of the current anomaly. The specific calculation method is as follows: If the current inspection cycle triggers the three-phase deviation judgment, the phase deviation intensity value is obtained by proportional conversion and normalization based on the maximum phase temperature difference and the preset phase temperature difference threshold. The value ranges from 0 to 1, and the larger the value, the more severe the three-phase temperature imbalance.

[0052] If the current inspection cycle triggers adaptive deviation judgment, the adaptive deviation intensity value is obtained by proportionally converting and normalizing the adaptive deviation coefficient with a preset deviation coefficient benchmark. The preset deviation coefficient benchmark is set based on operational experience; in this embodiment, it is set to 0.5. Based on the deviation judgment results obtained within the current inspection cycle, the comprehensive deviation confidence level is determined using the following rules: Scenario 1: When both the phase deviation intensity value and the adaptive deviation intensity value are obtained simultaneously, a weighted fusion calculation is performed according to the preset fusion rules. That is, the product of the phase deviation intensity value and its weight coefficient is added to the product of the adaptive deviation intensity value and its weight coefficient. The sum of the weight coefficients is 1. In this embodiment, the weight coefficient corresponding to the phase deviation intensity value is 0.6, and the weight coefficient corresponding to the adaptive deviation intensity value is 0.4, which reflects the fusion strategy of taking the three-phase deviation as the main factor and the adaptive deviation as the auxiliary factor.

[0053] Case 2: When only the phase deviation intensity value is obtained, this value is used as the overall deviation confidence level.

[0054] Scenario 3: When only the adaptive deviation strength value is obtained, this value is used as the overall deviation confidence level.

[0055] Scenario 4: If one or more of the following are simultaneously obtained: absolute over-temperature deviation flag, abnormal heating rate deviation flag, or abnormal temperature gradient deviation flag, then for each type of triggered deviation, define its deviation intensity value s. i Its specific expression is: ; In the formula, 'a' represents the parameter corresponding to the current deviation mark, namely the temperature trend value corresponding to the absolute over-temperature deviation, the temperature rate corresponding to the abnormal temperature rise rate deviation, or the temperature gradient corresponding to the abnormal temperature gradient deviation. 'a0' is the corresponding preset value, which is the average value of historical values ​​during the historical inspection process. A benchmark weight is preset for each deviation type to characterize the relative importance of this type of abnormality in fault diagnosis. The sum of all benchmark weights is 1, and the specific value is adjusted by the preset personnel according to the current working conditions.

[0056] Let S be the set of deviation types actually triggered in the current inspection cycle. Based on the number of elements in set S, a unified fusion rule is used to calculate the comprehensive deviation confidence: the comprehensive deviation confidence is the weighted sum of the intensity values ​​of each type of deviation in set S and their corresponding weights.

[0057] If the overall deviation confidence level is within the first overall deviation interval, the current anomaly is determined to be highly credible. The overall deviation flag set for the current inspection cycle is immediately output, and a complete alarm response is triggered. This includes uploading the overall deviation flag set, containing all triggered deviation types, deviation magnitudes, corresponding phases and locations, and timestamps, to the main station; capturing the dual-light fusion image and visible light close-up image at the moment of the anomaly, and uploading them together with the alarm information; and pushing alarm notifications to pre-set maintenance personnel, prompting them to conduct on-site inspections of the cable terminals.

[0058] If the overall deviation confidence level is within the second overall deviation interval, the corresponding monitoring area will be marked as an abnormal state pending confirmation. Based on the current overall deviation confidence level, the corresponding sampling frequency increase value will be matched in the preset sampling frequency mapping table. This sampling frequency increase value will be used as the sampling frequency for the next inspection cycle. The deviation judgment result of the next inspection cycle will be waited for and compared with the current result. The validity of the fault terminal will be re-determined. The sampling frequency mapping table predefines the sampling frequency multiples corresponding to different confidence level intervals.

[0059] If the overall deviation confidence level is within the third overall deviation interval, the current deviation is determined to be a low-confidence anomaly, and a low-confidence anomaly warning is sent.

[0060] To more clearly illustrate the processing logic when multiple deviation markers are combined, two typical examples are given below. The first comprehensive deviation interval is defined as [0.8, 1], the second comprehensive deviation interval is defined as [0.5, 0.8), and the third comprehensive deviation interval is defined as [0, 0.5]. Furthermore, the sampling rate is refined within the second comprehensive deviation interval by dividing the second comprehensive deviation interval into three equal parts, which correspond to a sampling frequency increase of 2 times, 3 times, and 4 times, respectively.

[0061] Example 1: Simultaneous Triggering of Interphase Deviation and Adaptive Deviation: Suppose that during the current inspection cycle, a certain monitoring area simultaneously triggers interphase deviation and adaptive deviation, and the calculated interphase deviation intensity value is 0.9, the adaptive deviation intensity value is 0.7, and the comprehensive deviation confidence level is calculated to be 0.82 according to the fusion rule, which is within the first comprehensive deviation interval. The system determines it to be a highly reliable anomaly, immediately outputs a comprehensive deviation flag set containing interphase deviation flags and adaptive deviation flags, and triggers a complete alarm response.

[0062] Example 2: Simultaneous triggering of phase deviation, adaptive deviation, and abnormal heating rate: Suppose that a certain monitoring area simultaneously triggers phase deviation with a value of 0.8, adaptive deviation with a value of 0.6, and abnormal heating rate with a value of 0.9. The weights of the three are set to 0.5, 0.3, and 0.2, respectively. Then the comprehensive deviation confidence level is 0.76, which is in the second comprehensive deviation interval. This area is marked as an abnormal state pending confirmation. Based on 0.76, the corresponding sampling frequency increase value is matched in the sampling frequency mapping table. Intensified sampling is carried out in the next inspection cycle. A final judgment is made after further confirmation.

[0063] Taking an outdoor terminal of a 220kV cable line in a coastal city, which has been in operation for 5 years, as an example, the line's load exhibits a typical daytime high and nighttime low characteristic: the peak daytime load is approximately 480A, and the off-peak nighttime load is approximately 120A. The local climate is humid, with frequent rain in summer and fog in winter, resulting in complex environmental conditions.

[0064] Dual-light fusion images of the cable terminal area were acquired, and monitoring areas were divided based on pre-calibrated template images, including the conductor connection areas of phases A, B, and C, the outer surface area of ​​the insulation, and the outer surface area of ​​the bushing. The temperature trend values ​​of each monitoring area were within the normal fluctuation range, and no deviation judgment was triggered.

[0065] At 2:00 AM on the 35th day, the system acquired a set of dual-light fusion images. At this time, the load current was a low-load of 125A, the ambient temperature was 22℃, and the humidity was 75%. The temperature trend value of the A-phase conductor connection area was 31.2℃, phase B was 30.1℃, and phase C was 30.3℃. A historical temperature trend value sequence of 720 data points from the past 30 days for the A-phase conductor connection area was retrieved, and the calculated statistical mean was 28.5℃, with a standard deviation of 0.8℃. Based on the preset confidence coefficient of 3, an adaptive reference interval of 26.1-30.9℃ was constructed, triggering adaptive deviation. The deviation type was recorded as high adaptive over-limit, and the deviation magnitude was 0.3℃.

[0066] Based on the adaptive deviation coefficient and the preset deviation coefficient benchmark, the adaptive deviation intensity value is calculated to be 0.8. Currently, only adaptive deviation is triggered, so the comprehensive deviation confidence level is 0.8. It is in the first comprehensive deviation interval and is judged as a highly reliable anomaly. An alarm response is immediately triggered, the comprehensive deviation flag set containing the adaptive deviation flag is output, and a warning message is pushed to the operation and maintenance personnel: an early temperature anomaly has occurred in the A phase conductor connection area. It is recommended to pay close attention.

[0067] like Figure 5The diagram shown is a structural schematic of a cable terminal fault diagnosis system based on infrared online monitoring provided in an embodiment of the present invention. The system includes the following modules: a dual-light fusion module, used to acquire visible light images and infrared thermal images of the cable terminal area and overlay them to generate a dual-light fusion image; a monitoring area division module, used to divide at least one monitoring area on the dual-light fusion image based on pre-calibrated components of the cable terminal area, obtaining a set of monitoring areas for each component of the cable terminal and its component labels; a dynamic trend filtering module, used to extract temperature characteristic parameters of the monitoring area and perform dynamic trend filtering on the temperature characteristic parameters to obtain temperature trend values; and a deviation comparison and judgment module, used to perform deviation comparison and judgment on the obtained temperature trend values, thereby generating a comprehensive deviation marker set for determining the validity of the terminal fault.

[0068] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0069] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0070] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0071] Those skilled in the art will recognize that the units 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 implementations should not be considered beyond the scope of this invention.

[0072] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0073] In the embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0074] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0075] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0076] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0077] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A cable terminal fault diagnosis method based on infrared online monitoring, characterized in that, Includes the following steps: S1: Acquire visible light and infrared thermal images of the cable terminal area and overlay the images to generate a dual-light fused image; S2, based on the pre-calibrated components of the cable terminal area, divide at least one monitoring area on the dual-light fusion image to obtain the monitoring area set of each component of the cable terminal and its component label; The monitoring area includes at least one or more of the following: conductor connection or crimping point area, insulation outer surface area, and terminal head or sleeve outer surface area; S3, extract the temperature characteristic parameters of the monitoring area, and perform dynamic trend filtering on the temperature characteristic parameters to obtain temperature trend values; The temperature characteristic parameters include at least one or more of the following: the highest temperature of the monitoring area, the average temperature of the monitoring area, the heating rate and direction, and the temperature trend value is used to characterize the stable change direction and magnitude of temperature over time in the monitoring area. S4, the obtained temperature trend value is compared and judged to generate a comprehensive deviation mark set for terminal fault validity determination; The comprehensive deviation marker set includes at least one or more of the following: absolute over-temperature deviation marker, relative temperature rise deviation marker, abnormal temperature rise rate deviation marker, and abnormal temperature gradient deviation marker.

2. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 1, characterized in that, The image overlay process is as follows: At the same acquisition time, the visible light raw image and the infrared raw image of the cable terminal area are acquired simultaneously; Scale-invariant feature transformations are performed on the original visible light image and the original infrared image respectively to obtain a set of visible light feature points and a set of infrared feature points. Based on the feature descriptors, bidirectional matching is performed on the set of visible light feature points and the set of infrared feature points. The bidirectional matching includes: calculating the descriptor similarity between any visible light feature point and a candidate infrared feature point, and simultaneously verifying the reverse optimal matching relationship between the candidate infrared feature point and the visible light feature point to obtain candidate matching point pairs. Candidate matching point pairs with descriptor similarity greater than the preset descriptor similarity are retained to form an initial matching point set, and spatial transformation relationships are constructed.

3. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 2, characterized in that, The specific process for constructing the spatial transformation relationship is as follows: Using the coordinates of visible light feature points and corresponding infrared feature points in the initial set of matching points as constraint samples, a perspective transformation model between the visible light image coordinate system and the infrared image coordinate system is established to obtain the spatial transformation matrix to be solved. The parameters of the spatial transformation matrix are solved to minimize the sum of squared errors between each visible light feature point and its corresponding infrared feature point after mapping by the spatial transformation matrix, so as to obtain the matrix element values ​​of the spatial transformation matrix. The pixel coordinates of the original infrared image are transformed according to the spatial transformation matrix to determine the corresponding position of each infrared pixel in the visible light image coordinate system. Interpolation and resampling are performed on the infrared pixels after coordinate transformation to generate a transformed infrared image that is the same size as the original visible light image and spatially aligned. The transformed infrared image is superimposed onto the corresponding area of ​​the original visible light image according to a preset overlay method to obtain a dual-light fusion image, and the monitoring area is divided based on the dual-light fusion image.

4. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 3, characterized in that, The specific process for dividing the monitoring area is as follows: Based on the pre-calibrated template image, the pixel coordinate range of each component in the cable terminal area in the dual-light fusion image is determined, and the pixel coordinate range is mapped one-to-one with the corresponding component label to generate a component coordinate mapping table; Based on the component coordinate mapping table, the pixel coordinate range corresponding to each component is read on the dual-light fusion image to form an initial monitoring area set that corresponds one-to-one with the component label, so as to carry out hierarchical monitoring.

5. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 4, characterized in that, The specific process of the hierarchical monitoring is as follows: Obtain the fault frequency coefficient and temperature fluctuation coefficient of each initial monitoring area, and construct a regional fluctuation index based on the fault frequency coefficient and temperature fluctuation coefficient to quantify the fault probability of each monitoring area. The fault frequency coefficient is obtained by normalizing the statistical value of the number of times the corresponding component in the current initial monitoring area has experienced temperature abnormality alarms during the historical operating cycle. The temperature fluctuation coefficient is determined based on the ratio of the standard deviation of the temperature to the preset fluctuation benchmark value within a preset number of inspection cycles in the current initial monitoring area. If the regional fluctuation index is greater than the preset regional fluctuation index, the corresponding initial monitoring area is recorded as the first deviation level area, and the temperature feature parameters are repeatedly extracted within the inspection cycle at the first sampling frequency, and repeated trend filtering is performed to obtain the first temperature trend value sequence. Conversely, the corresponding initial monitoring area is recorded as the second deviation level area, and the temperature characteristic parameters are extracted once within the inspection cycle at the second sampling frequency, and a single trend filter is performed to obtain the second temperature trend value. The first sampling frequency is greater than the second sampling frequency, and the time sampling density of the first temperature trend value sequence is greater than the time sampling density of the second temperature trend value.

6. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 1, characterized in that, The deviation comparison judgment process is as follows: Within the current inspection cycle, the current temperature trend value of each phase terminal of the cable at the same location monitoring area at the same sampling time is obtained. Each phase terminal of the cable refers to the terminal head or conductor connection end of different phases in the same circuit cable, including the first phase terminal, the second phase terminal and the third phase terminal. The current temperature trend values ​​corresponding to the first phase terminal, the second phase terminal, and the third phase terminal are respectively recorded as the first phase temperature trend value, the second phase temperature trend value, and the third phase temperature trend value, and the absolute difference between each pair of the three is calculated. The maximum value of the absolute differences between any two phases is determined as the maximum temperature difference between phases, and the phase deviation coefficient is determined based on the ratio between the maximum temperature difference between phases and the preset reference temperature difference value. This coefficient is used to characterize the degree of temperature imbalance in the monitoring area of ​​the same location at each phase terminal of the cable. When the maximum temperature difference between the phases exceeds the preset temperature difference threshold between the phases, it is determined that a phase deviation has been triggered. Based on the relationship between the temperature trend values ​​of the two phases that contribute to the formation of the maximum temperature difference between the two phases, the phase with the higher temperature trend value is identified as the abnormal phase. A phase deviation mark is generated for the abnormal phase and the deviation magnitude is recorded. When the maximum temperature difference between the phases is not greater than the preset temperature difference threshold between the phases, phase deviation will not be triggered.

7. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 6, characterized in that, The deviation comparison judgment also includes: For any monitoring area, the historical temperature trend value sequence of the monitoring area within a preset historical window is obtained from the historical database, and statistical analysis is performed to obtain the statistical mean and statistical standard deviation. The statistical standard deviation is proportional to the historical temperature standard deviation to obtain an adaptive deviation coefficient; Based on the statistical mean, the statistical standard deviation, and the preset confidence coefficient, an adaptive reference interval for the monitoring area is constructed. Its upper and lower limits are respectively: the statistical mean plus the standard deviation of the preset confidence coefficient and the adaptive deviation coefficient, and the statistical mean minus the standard deviation of the preset confidence coefficient and the adaptive deviation coefficient. Obtain the temperature trend value of the monitored area during the current inspection cycle; If the temperature trend value is greater than the upper limit of the adaptive reference range, an adaptive deviation is triggered, and the deviation type is recorded as high adaptive overlimit and its corresponding deviation magnitude. If the temperature trend value is less than the lower limit of the adaptive reference range, an adaptive deviation is triggered, and the deviation type is recorded as low adaptive overlimit and its corresponding deviation magnitude. If the temperature trend value is within the adaptive reference range, adaptive deviation will not be triggered.

8. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 1, characterized in that, The specific process for determining the validity of terminal faults is as follows: Based on the maximum interphase temperature difference or adaptive deviation coefficient obtained within the current inspection cycle, the comprehensive deviation confidence level of the comprehensive deviation marker set in the current inspection cycle is determined, specifically as follows: The maximum temperature difference between phases is proportionally converted to a preset temperature difference threshold between phases and then normalized to obtain the phase deviation intensity value. The adaptive deviation coefficient is proportionally converted and normalized with the preset deviation coefficient benchmark to obtain the adaptive deviation intensity value. When both the phase deviation strength value and the adaptive deviation strength value are obtained simultaneously, the two are fused according to the preset fusion rule to obtain the comprehensive deviation confidence. When only one of them is obtained, the deviation strength value is used as the overall deviation confidence level.

9. The cable terminal fault diagnosis method based on infrared online monitoring as described in claim 8, characterized in that, The determination of the validity of the terminal fault also includes: If the overall deviation confidence level is within the first overall deviation interval, the current overall deviation flag set is output and an alarm response is triggered to prompt the preset personnel to inspect the cable terminal. If the overall deviation confidence level is within the second overall deviation interval, the corresponding monitoring area is marked as an abnormal state to be confirmed. The corresponding sampling frequency increase value is matched in the preset sampling frequency mapping table according to the current overall deviation confidence level. The sampling frequency increase value is used as the sampling frequency of the next inspection cycle. The deviation judgment result of the next inspection cycle is waited for and compared with the current result to re-determine the validity of the fault terminal. If the overall deviation confidence level is within the third overall deviation interval, the current deviation is determined to be a low-confidence anomaly, and a low-confidence anomaly warning is sent.

10. A cable terminal fault diagnosis system based on infrared online monitoring, employing the cable terminal fault diagnosis method based on infrared online monitoring as described in any one of claims 1-9, comprising the following modules: The dual-light fusion module is used to acquire visible light images and infrared thermal images of the cable terminal area and overlay the images to generate a dual-light fused image. The monitoring area division module is used to divide at least one monitoring area on the dual-light fusion image based on the pre-calibrated components of the cable terminal area, so as to obtain the monitoring area set of each component of the cable terminal and its component label. The dynamic trend filtering module is used to extract the temperature characteristic parameters of the monitoring area and perform dynamic trend filtering on the temperature characteristic parameters to obtain temperature trend values. The deviation comparison and judgment module is used to compare and judge the deviation of the acquired temperature trend value, thereby generating a comprehensive deviation mark set for determining the validity of terminal faults.

Images