A method, system, device and medium for evaluating the operating condition of an overhead line external insulation device of a distribution network

Abstract

The application discloses a kind of distribution network overhead line external insulation equipment operating condition evaluation method, system, equipment and medium, comprising: obtaining target insulation equipment surface image and identifying new color covering, if there is, generate time stamp containing contamination event record and set to strengthen monitoring mode;When in strengthening monitoring mode, real-time acquisition leakage current and partial discharge intensity, any parameter is over threshold value, then obtain the dynamic response data in the first natural humidity change period after contamination event record;Dynamic response data is compared with benchmark response characteristics, and attribution judgment result is generated by fusing contamination characteristic information;Attribution judgment result and dynamic response data are stored in digital twin archives;Based on digital twin archives, extract feature vector and carry out similarity search with cloud case base, update attribution judgment weight parameter, obtain external insulation equipment operating condition evaluation report.The application effectively improves the accuracy of fault diagnosis and the intelligent level of equipment operation and maintenance.

Description

Technical Field

[0001] This invention relates to the field of overhead line condition monitoring technology in power distribution networks, and in particular to a method, system, equipment, and medium for assessing the operating status of external insulation equipment of overhead lines in power distribution networks. Background Technology

[0002] The external insulation equipment of overhead distribution lines is exposed to the outdoor environment for extended periods, making it susceptible to surface contamination such as bird droppings, as well as latent defects like internal cracks caused by long-term operation. Currently, the operational status assessment of distribution network external insulation equipment mainly relies on manual inspections or single online monitoring methods. For example, drones are used to capture images of insulator surfaces to identify obvious contamination or damage, or leakage current sensors are installed to monitor electrical anomalies. However, these methods are often independent; image data only reflects surface conditions, and electrical data only reflects the current operating status, lacking a comprehensive assessment mechanism that effectively integrates both.

[0003] When bird droppings cover the surface of insulating equipment, if there happens to be an internal crack in that area, the droppings, as conductive contaminants, can trigger electrical anomalies such as increased leakage current or partial discharge. However, existing single monitoring methods struggle to distinguish whether these anomalies are simply caused by surface bird droppings or by internal cracks beneath the droppings. This ambiguity in attribution prevents maintenance personnel from taking precise action—cleaning it as bird droppings might overlook the underlying cracks; replacing it due to cracks could result in unnecessary resource waste and power outage losses. Therefore, accurately identifying the root cause of equipment defects by penetrating surface contamination has become a pressing technical challenge in the condition assessment of external insulation equipment in distribution networks. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method, system, equipment, and medium for assessing the operational status of external insulation equipment of overhead distribution lines, which solves the problems of existing assessment methods for external insulation equipment of distribution networks being independent of each other and lacking a multi-physical quantity fusion mechanism, leading to fuzzy attribution and difficulties in operation and maintenance decisions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for evaluating the operational status of external insulation equipment of overhead distribution lines, comprising: Acquire a surface image of the target insulation device and identify whether there is a new cover of a different color. In response to the identification of a new cover of a different color, generate a dirt event record with a timestamp and set the target insulation device to enhanced monitoring mode. In response to the target insulation device being in enhanced monitoring mode, the leakage current and partial discharge intensity of the target insulation device are collected in real time; in response to the detection that either the leakage current or the partial discharge intensity exceeds a preset threshold, the dynamic response data of the target insulation device within the first natural humidity change cycle after the pollution event is recorded are obtained. The dynamic response data is compared with the pre-stored baseline response features, and the filth feature information in the filth event record is fused to generate an attribution judgment result for the current abnormal signal. The attribution judgment results and the corresponding dynamic response data are stored in the digital twin file of the target insulation device; Feature vectors are extracted from the digital twin archives and similarity searches are performed with the cloud case library. Attribution judgment weight parameters are updated based on the search results to obtain an assessment report on the operating status of external insulation equipment.

[0007] As a preferred embodiment of the method for evaluating the operational status of external insulation equipment of overhead distribution lines according to the present invention, the step of identifying whether there is a newly added discolored covering includes: Collect the current surface image of the target insulation device at the current time point according to a preset cycle, and obtain historical surface images from the same viewpoint; Image registration and illumination correction are performed on the current surface image and the historical surface image to generate aligned and standardized current image and standardized historical image; The standardized current image and the standardized historical image are subjected to differential processing to extract the difference region; Extract the geometric and color features of the differential region, compare the geometric and color features with a preset dirt feature threshold range, and determine that the differential region is a newly added heterochromatic covering when the geometric and color features meet the dirt feature threshold range.

[0008] As a preferred embodiment of the method for evaluating the operational status of external insulation equipment of overhead distribution lines according to the present invention, wherein setting the target insulation equipment to an enhanced monitoring mode includes: When the differential area is determined to be a newly added color cover, the current time point is recorded as the time of contamination discovery, and the device code and geographical coordinates of the target insulation device are obtained to generate a contamination event record containing the time of contamination discovery, the device code, and the geographical coordinates. The pollution event is recorded in the operation and maintenance file of the target insulation equipment, and the monitoring status identifier of the target insulation equipment is changed from routine monitoring to enhanced monitoring. The enhanced monitoring mode includes increasing the data acquisition frequency of leakage current and partial discharge intensity of the target insulation equipment, and extending the validity period of the pollution event records.

[0009] As a preferred embodiment of the method for evaluating the operational status of external insulation equipment of overhead distribution lines according to the present invention, the step of obtaining dynamic response data of the target insulation equipment within the first natural humidity change cycle after the pollution event is recorded includes: When the target insulation equipment is in enhanced monitoring mode, the leakage current amplitude and partial discharge pulse frequency of the target insulation equipment are continuously collected, and the ambient relative humidity data are collected simultaneously. In response to the leakage current amplitude exceeding a preset current threshold, a dynamic response data capture process is triggered. In response to the partial discharge pulse frequency exceeding a preset frequency threshold, a dynamic response data capture process is triggered; Starting from the triggering moment, record the process data of the leakage current amplitude and the partial discharge pulse frequency changing with the ambient relative humidity during the first natural humidity change cycle after the pollution event is recorded.

[0010] The beneficial effects of this preferred technical solution are that, through the dual threshold triggering mechanism of linked leakage current and partial discharge, the dynamic evolution of electrical parameters within the first complete natural humidity change cycle after a contamination event is accurately locked, effectively eliminating random noise interference and completely reproducing the conductivity mechanism of the contamination layer under the influence of humidity, thereby significantly improving the accuracy and reliability of abnormal signal attribution analysis.

[0011] As a preferred embodiment of the method for evaluating the operational status of external insulation equipment of overhead distribution lines according to the present invention, the triggering dynamic response data capture process includes: The real-time collected leakage current amplitude is compared with the preset current threshold, and the partial discharge pulse frequency is compared with the preset frequency threshold. When any monitored value exceeds the corresponding threshold, a trigger signal containing the type of the out-of-limit parameter and the time of the out-of-limit is generated; In response to the trigger signal, leakage current data, partial discharge data and synchronous ambient relative humidity data within a preset time period before and after the trigger time are solidified and stored to generate dynamic response data fragments to be analyzed. Simultaneously, natural humidity change cycle monitoring is initiated to identify the first natural humidity change cycle after the aforementioned pollution event is recorded.

[0012] As a preferred embodiment of the method for evaluating the operating status of external insulation equipment of overhead distribution lines according to the present invention, the generation of attribution judgment results for the current abnormal signal includes: Response feature parameters are extracted from the dynamic response data, including response delay time, response peak value, and recovery time. The response feature parameters are compared with the pre-stored benchmark response features to obtain the feature deviation vector; Obtain the contamination event records of the target insulation equipment, and extract the identification information of whether there are any newly added different colored coverings in the contamination event records; The feature deviation vector and the identification information are input into the attribution discrimination model, and the attribution discrimination model outputs the attribution judgment result based on the magnitude of the feature deviation vector and the value of the identification information.

[0013] As a preferred embodiment of the method for evaluating the operational status of external insulation equipment of overhead distribution lines according to the present invention, the step of updating the attribution judgment weight parameters based on the search results includes: Feature vectors are extracted from the dynamic response data. The feature vectors consist of response delay time deviation, response peak deviation, and recovery time deviation. The similarity between the feature vector and the feature vectors of multiple historical cases in the cloud case library is calculated to obtain similar historical cases whose similarity exceeds a preset similarity threshold. Obtain the historical attribution judgment results and historical maintenance conclusions corresponding to the similar historical cases; The historical attribution judgment results and the historical maintenance conclusions are compared with the currently generated attribution judgment results; Based on the comparison results, the weight parameters used for subsequent attribution judgment in the attribution discrimination model are updated.

[0014] The beneficial effects of this preferred technical solution are that by extracting dynamic response feature vectors and matching them with historical cases in the cloud, and by using historical attribution results verified by actual maintenance conclusions to perform closed-loop correction of the current model weights, adaptive iterative optimization of the attribution discrimination model is achieved, thereby significantly improving the accuracy and generalization ability of subsequent abnormal signal attribution judgment.

[0015] Secondly, the present invention provides a system for evaluating the operational status of external insulation equipment of overhead distribution lines, comprising: A dirt visual recognition module is used to acquire a surface image of the target insulation equipment and identify whether there is a new cover of a different color. In response to the identification of a new cover of a different color, a dirt event record containing a timestamp is generated, and the target insulation equipment is set to enhanced monitoring mode. The dynamic data capture module is used to collect the leakage current and partial discharge intensity of the target insulation device in real time in response to the target insulation device being in enhanced monitoring mode; and to acquire the dynamic response data of the target insulation device in the first natural humidity change cycle after the pollution event is recorded when either the leakage current or the partial discharge intensity is detected to exceed a preset threshold. The anomaly attribution analysis module is used to compare the dynamic response data with the pre-stored baseline response features and integrate the filth feature information in the filth event record to generate an attribution judgment result for the current anomaly signal. The data archiving module is used to store the attribution judgment results and the corresponding dynamic response data into the digital twin file of the target insulation device; The evaluation and optimization module is used to extract feature vectors based on the digital twin archives and perform similarity retrieval with the cloud case library. Based on the retrieval results, the attribution judgment weight parameters are updated to obtain an evaluation report on the operating status of the external insulation equipment.

[0016] Thirdly, the present invention provides an electronic device, comprising: Memory, used to store programs; A processor is configured to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method for evaluating the operating status of external insulation equipment of overhead distribution lines.

[0017] Fourthly, the present invention provides a computer-readable storage medium, comprising: when the program is executed by a processor, the steps of implementing the method for evaluating the operating status of external insulation equipment of overhead distribution lines.

[0018] The beneficial effects of this invention are as follows: By constructing a trigger mechanism based on the identification of new discolored coverings in surface images and linking it to switch enhanced monitoring modes, this invention achieves early and accurate perception of the contamination status of insulating equipment and dynamic adaptive adjustment of monitoring strategies, avoiding the lag of traditional fixed-cycle monitoring. By locking the leakage current and partial discharge dynamic response data within the first natural humidity change cycle after a contamination event and fusing and comparing them with pre-stored benchmark response features, it achieves in-depth tracing of the physical causes of abnormal signals under real-world humidity alternation conditions, effectively eliminating random noise interference and improving the confidence of attribution judgment. By establishing a digital twin archive and retrieving and matching real-time feature vectors with a cloud-based case library to update attribution weight parameters in reverse, it achieves closed-loop self-evolution of the model based on historical maintenance conclusion verification, significantly improving the generalization ability and long-term prediction accuracy of the operation status assessment report in different scenarios. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the 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. Wherein: Figure 1 This is a basic flowchart illustrating a method for evaluating the operational status of external insulation equipment of overhead distribution lines, as provided in one embodiment of the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0021] Example 1, referring to Figure 1 As an embodiment of the present invention, a method for evaluating the operating status of external insulation equipment of overhead distribution lines is provided, comprising: S100: Acquire a surface image of the target insulation device and identify whether there is a new colored covering. In response to the identification of a new colored covering, generate a dirt event record with a timestamp and set the target insulation device to enhanced monitoring mode. S200: In response to the target insulation equipment being in enhanced monitoring mode, the leakage current and partial discharge intensity of the target insulation equipment are collected in real time; in response to the detection that either the leakage current or the partial discharge intensity exceeds a preset threshold, the dynamic response data of the target insulation equipment within the first natural humidity change cycle after the pollution event is recorded are obtained. S300: Compares the dynamic response data with the pre-stored baseline response features and integrates the filth feature information in the filth event record to generate an attribution judgment result for the current abnormal signal. S400: Store the attribution judgment results and the corresponding dynamic response data into the digital twin file of the target insulation device; S500: Based on digital twin archives, feature vectors are extracted and similarity searches are performed with cloud-based case databases. The attribution judgment weight parameters are updated according to the search results to obtain an assessment report on the operating status of external insulation equipment.

[0022] It should be noted that existing assessment methods for external insulation equipment in distribution networks face severe challenges during operation. These include: traditional manual inspections being limited by geographical environment and weather conditions, resulting in long cycles, numerous blind spots, and difficulty in capturing sudden pollution accumulation in real time, leading to a significant lag in condition perception; conventional online monitoring relies heavily on single leakage current or meteorological data, lacking comprehensive analysis of the coupling characteristics of multiple physical quantities such as partial discharge, making it difficult to accurately distinguish between normal fluctuations and early fault signs in complex and variable humidity and pollution interaction environments, resulting in high false alarm and false negative rates; existing assessment models are mostly based on static thresholds or general empirical formulas, lacking adaptive learning capabilities for specific equipment's historical operating conditions, and failing to effectively quantify the nonlinear conductivity mechanism of the pollution layer under dynamic humidity changes, leading to ambiguous defect attribution and large deviations in remaining life prediction; due to the lack of a closed-loop feedback mechanism of digital twins and cloud-based case libraries, historical operation and maintenance data has not been transformed into a driving force for model optimization, making it difficult for the assessment system to iterate and upgrade with the evolution of the power grid environment, and unable to support refined management and proactive defense of distribution network insulation status.

[0023] Therefore, addressing the aforementioned issues of existing independent assessment methods for external insulation equipment in distribution networks, lack of multi-physical quantity fusion mechanisms leading to ambiguous attribution and difficulties in operation and maintenance decisions, this paper achieves a breakthrough in the entire chain of technology for external insulation equipment in distribution networks. This breakthrough is achieved through steps S100-S400, including constructing an enhanced monitoring mode based on image recognition triggering, locking dynamic response data of multiple physical quantities within the first natural humidity change cycle, and utilizing a collaborative mechanism for closed-loop updating of weight parameters using a cloud-based historical case library. This breakthrough significantly improves the real-time performance, accuracy, and generalization ability of condition assessment.

[0024] Example 2, this is an embodiment of the present invention, which provides a method for evaluating the operating status of external insulation equipment of overhead distribution lines based on the previous embodiment, including: In this embodiment of the application, step S100, identifying whether there is a newly added cover of a different color, includes: Collect current surface images of the target insulation equipment at the current time point according to a preset cycle, and acquire historical surface images from the same perspective; Perform image registration and illumination correction on the current surface image and historical surface images to generate aligned and standardized current and historical images; The current standardized image is compared with the historical standardized image using a difference processing method to extract the difference region; Extract the geometric and color features of the differential region, compare the geometric and color features with the preset dirt feature threshold range, and determine the differential region as a newly added heterochromatic covering when the geometric and color features meet the dirt feature threshold range.

[0025] In this embodiment, step S100 uses a high-definition camera mounted on a drone to photograph the target insulation equipment according to a preset inspection cycle, such as once a week, to obtain a current surface image including a timestamp. Simultaneously, historical surface images taken from the same viewpoint at the previous time point are retrieved from a historical image database. Since the drone's hovering position, shooting angle, and ambient lighting may differ between the two shots, the two images need to be preprocessed to ensure comparability. Specifically, image registration is performed between the current surface image and the historical surface image. The purpose of image registration is to align the shooting perspectives of the two images.

[0026] In this embodiment, step S100 employs a registration method based on feature point matching. Scale-invariant feature transformation points are extracted from the two images. Matching feature point pairs are selected using the Random Sample Consensus Algorithm (RANSAC), and the perspective transformation matrix between the two images is calculated. The current surface image is then transformed according to this perspective transformation matrix to ensure its viewpoint aligns with the historical surface image. A histogram matching method is used to calculate the grayscale histograms of both images. Using the histogram of the historical surface image as a reference, the registered current surface image undergoes grayscale transformation to make the brightness distribution of the two images more consistent. After the above processing, an aligned, standardized current image and a standardized historical image are generated.

[0027] In this embodiment, step S100 involves performing differential processing on the standardized current image and the standardized historical image. The difference in grayscale values ​​between the two images is calculated pixel-by-pixel to generate a differential image. Pixels with significant grayscale value changes in the differential image represent the regions where the two images differ. The differential image is then binarized, with a grayscale difference threshold of 30 (ranging from 0 to 255). Pixels with grayscale differences exceeding this threshold are marked as foreground points, and the rest as background points. A morphological closing operation is performed on the binarized image to connect adjacent foreground points, forming several connected candidate regions, thus extracting the differential regions.

[0028] In an optional implementation, the surface image difference recognition and preprocessing method in step S100 can also directly construct and train a deep convolutional neural network (such as U-Net), input the surface image collected by the current UAV, and the model automatically learns and outputs a pixel-level mask of dirt to identify the different colored coverings.

[0029] In an optional implementation, the surface image difference identification and preprocessing method in step S100 can also replace the visible light camera carried by the UAV with a multispectral or hyperspectral imager to collect image data containing rich spectral information, construct spectral indices or perform spectral angle matching analysis using the differences in spectral reflectance characteristics of specific dirt in different bands, and directly identify discolored coverings by classifying spectral features.

[0030] In this embodiment, step S100 involves feature extraction and discrimination for each differential region. The extracted features include geometric features and color features. Geometric features include region area (number of pixels) and region aspect ratio (ratio of the longer to shorter side of the smallest bounding rectangle). Color features include the region's average hue and average saturation values ​​in the original image. The extracted features are compared with a preset dirt feature threshold range. The preset dirt feature threshold range is set based on statistical results of typical bird droppings samples. To accommodate imaging differences caused by different shooting distances, the region area threshold is preferably expressed as a relative area, i.e., the percentage of pixels in the differential region relative to the total number of pixels in the projected area of ​​the umbrella skirt of the target insulating device, with a threshold range of 1% to 30%. The aspect ratio threshold for geometric features is set to 1.5-8.0. The hue threshold for color features is set to 20-60, and the saturation threshold is set to 30-80. When all features of the differential region fall within the corresponding threshold range, the differential region is determined to be a newly added discolored covering, that is, newly attached bird droppings are identified on the surface of the insulating equipment. This realizes the automatic identification of newly added bird droppings on the surface of the insulating equipment and provides basic data for recording dirt events in subsequent steps.

[0031] In this embodiment of the application, step S100, setting the target insulation device to enhanced monitoring mode, includes: When the differential area is determined to be a newly added color cover, the current time point is recorded as the time of contamination discovery, and the equipment code and geographical coordinates of the target insulation equipment are obtained to generate a contamination event record containing the time of contamination discovery, equipment code and geographical coordinates. The pollution incident records are stored in the operation and maintenance file of the target insulation equipment, and the monitoring status indicator of the target insulation equipment is changed from routine monitoring to enhanced monitoring. Enhanced monitoring methods include increasing the frequency of data acquisition for leakage current and partial discharge intensity of target insulated equipment, and extending the validity period of pollution event records.

[0032] In this embodiment, step S200 records the current system time as the pollution detection time. Simultaneously, the system retrieves the device code and geographic coordinates of the target insulation device from the device management database. The device code is a unique identifier assigned to each insulation device in the power distribution network asset management system. The geographic coordinates are obtained through GPS information recorded by a drone or preset coordinate information from a fixed camera terminal. The system combines the pollution detection time, device code, and geographic coordinates to generate a complete pollution event record. The data structure of this record is as follows: Contamination event log = {equipment code, geographic coordinates, time of contamination discovery}; In this embodiment of the application, in step S200, the system records the pollution event into the operation and maintenance file of the target insulation equipment. The operation and maintenance file is a collection of historical data of the equipment stored in the background server, which is used to record various events and monitoring data throughout the entire life cycle of the equipment. After the storage operation is completed, the system reads the current monitoring status identifier of the target insulation equipment. The monitoring status identifier is a parameter used to indicate the current monitoring level of the equipment. The initial state is normal monitoring. The system modifies the value of the identifier from normal monitoring to enhanced monitoring and writes the modified status identifier back to the equipment file.

[0033] In this embodiment of the application, the specific parameter adjustments in step S200 under enhanced monitoring status are as follows: (1) Increase the data acquisition frequency. Under normal monitoring conditions, the data acquisition frequency for leakage current and partial discharge intensity is once every 15 minutes. When the equipment enters enhanced monitoring mode, the acquisition frequency is adjusted to once every 1 minute. This adjustment is achieved by modifying the acquisition cycle configuration parameter in the field monitoring terminal; (2) Extend the validity period of contamination event records. The validity period of a contamination event record refers to the length of time that the record is used by the system as valid background information in attribution judgment. Under normal circumstances, the validity period of a contamination event record is set to 7 days. When the equipment enters enhanced monitoring status, the validity period is extended to 30 days. The adjustment of the validity period parameter is achieved by modifying the data processing rules of the backend server.

[0034] In this embodiment of the application, step S200, which involves locking and capturing dynamic response data within the first natural humidity change cycle after the timestamp, includes: When the target insulation equipment is in enhanced monitoring mode, the leakage current amplitude and partial discharge pulse frequency of the target insulation equipment are continuously collected, and the ambient relative humidity data are collected simultaneously. In response to the leakage current amplitude exceeding the preset current threshold, a dynamic response data capture process is triggered; In response to the partial discharge pulse frequency exceeding a preset frequency threshold, a dynamic response data acquisition process is triggered; Starting from the triggering moment, record the process data of leakage current amplitude and partial discharge pulse frequency as a function of ambient relative humidity during the first natural humidity change cycle after the pollution event is recorded.

[0035] In this embodiment of the application, step S200, which triggers the dynamic response data capture process, includes: The real-time collected leakage current amplitude is compared with the preset current threshold, and the partial discharge pulse frequency is compared with the preset frequency threshold. When any monitored value exceeds the corresponding threshold, a trigger signal containing the type of the out-of-limit parameter and the time of the out-of-limit is generated; In response to the trigger signal, leakage current data, partial discharge data and synchronous ambient relative humidity data within a preset time period before and after the trigger moment are solidified and stored to generate dynamic response data fragments to be analyzed. At the same time, natural humidity change cycle monitoring was initiated to identify the first natural humidity change cycle after a pollution event was recorded.

[0036] In this embodiment, during step S200, when the target insulation equipment is under enhanced monitoring, the on-site monitoring terminal continuously collects three data points at a frequency of once per minute: leakage current amplitude, partial discharge pulse frequency, and ambient relative humidity. The leakage current amplitude is obtained through a through-core current transformer installed on the grounding down conductor, and the unit is milliamperes (mA). The partial discharge pulse frequency is obtained by coupling the partial discharge signal through a high-frequency current transformer and counting the pulses, and the unit is pulses per second (Hz). The ambient relative humidity is obtained through a temperature and humidity sensor installed on the tower, and the unit is percentage RH. The collected data is transmitted to the backend server in real time. The server has two preset threshold parameters: a preset current threshold and a preset frequency threshold. In this embodiment, the preset current threshold is set to 5 mA, and the preset frequency threshold is set to 100 pulses per second. The server continuously compares the real-time received leakage current amplitude with the preset current threshold and the partial discharge pulse frequency with the preset frequency threshold.

[0037] In this embodiment, in step S200, when any monitored value exceeds the corresponding threshold, the system triggers a dynamic response data acquisition process. The specific steps are as follows: A trigger signal is generated. The trigger signal contains two pieces of information: the type of the over-limit parameter and the time of the over-limit. The type of the over-limit parameter identifies whether it is a leakage current over-limit or a partial discharge over-limit; if both are exceeded simultaneously, it is recorded as a dual-parameter over-limit. The time of the over-limit is precisely recorded, down to the second, at the Beijing time when the parameter first exceeds the threshold. In response to the trigger signal, the system stores the data within a preset time period before and after the trigger time.

[0038] In this embodiment, step S200 employs a ring-shaped buffer for temporary and permanent data storage. The ring-shaped buffer continuously stores all monitoring data from the most recent 60 minutes. When a trigger signal is received, the system retrieves a 60-minute data segment from the ring-shaped buffer, spanning from 30 minutes before to 30 minutes after the trigger moment, and transfers it from the temporary buffer to the permanent storage area, generating a dynamic response data segment to be analyzed. This data segment contains synchronously aligned leakage current data sequences, partial discharge data sequences, and ambient relative humidity data sequences, all strictly corresponding on the time axis.

[0039] In this embodiment of the application, in step S200, while triggering data solidification, the system initiates a natural humidity change cycle monitoring process to identify the first complete natural humidity change cycle that occurs after the pollution event is recorded. The natural humidity change cycle is defined as the complete process in which the ambient relative humidity rises continuously from a baseline value to a peak value, then falls back to the baseline value.

[0040] In this embodiment, the baseline value in step S200 is set to 1.2 times the lowest humidity of the day. The system continuously monitors the ambient relative humidity data. When the humidity value is detected to rise continuously by more than 3 percentage points and the final peak exceeds 80%RH, a humidity rise is determined to have begun. When the humidity value drops continuously from the peak by more than 3 percentage points and eventually returns to near the baseline value, a humidity fall is determined to have ended. The complete period from the start of the rise to the end of the fall constitutes a natural humidity change cycle.

[0041] In an optional implementation, the identification and definition of the natural humidity change cycle in step S200 can also be achieved by directly accessing the real-time weather forecast interface to obtain the humidity prediction sequence for future periods, dynamically defining the current humidity change stage based on the rising and falling trend and threshold range in the prediction curve, and using this as the input condition for subsequent insulation status assessment.

[0042] In an optional implementation, the identification and definition of the natural humidity change cycle in step S200 can also be achieved by collecting ambient temperature and relative humidity data in real time to calculate the dew point temperature, combining the surface temperature with the insulator surface heat conduction model, and dynamically defining the physical state stages of condensation and natural drying by judging whether the surface temperature is lower than the dew point temperature, thus serving as the basis for defining the humidity change cycle.

[0043] In this embodiment, in step S200, the system records the start time, peak time, and end time of the cycle, and associates the dynamic response data fragment stored in step S200 with the cycle. Only when the trigger time of the dynamic response data fragment is within the cycle, and the cycle occurs after the dirt event is recorded, is the data fragment confirmed as dynamic response data meeting the requirements of step S200, and used for subsequent attribution judgment. If the trigger time is not within any complete cycle, or if the cycle occurs before the dirt event is recorded, the system continues to wait and identify the next natural humidity change cycle. This enables accurate capture of the electrical response process data of the device under the excitation of natural humidity changes after enhanced monitoring of the device due to bird droppings pollution, providing a raw data basis for subsequent attribution judgment.

[0044] In this embodiment of the application, step S300, generating an attribution judgment result for the current abnormal signal, includes: Extract response feature parameters from dynamic response data. These response feature parameters include response latency, response peak value, and recovery time. The response feature parameters are compared with the pre-stored benchmark response features to obtain the feature deviation vector; Obtain the contamination event records of the target insulation equipment and extract the identification information of any newly added discolored coverings from the contamination event records; The feature bias vector and the identification information are input into the attribution discrimination model. The attribution discrimination model outputs the attribution judgment result based on the magnitude of the feature bias vector and the value of the identification information.

[0045] In this embodiment of the application, step S300 extracts characteristic parameters from the dynamic response data to characterize the device's response properties. The dynamic response data includes a sequence of leakage current amplitude, a sequence of partial discharge pulse frequencies, and a sequence of ambient relative humidity over a complete natural humidity variation cycle.

[0046] In this embodiment, step S300 extracts the following: response delay time, response peak value, and recovery time. The response delay time is defined as the time interval between the moment when the ambient relative humidity begins to rise and the moment when the leakage current amplitude begins to increase significantly. A sliding window algorithm is used to detect abrupt changes in the leakage current: the first-order difference of the leakage current sequence is calculated; when three consecutive difference values ​​exceed three times the historical average, it is determined that a significant increase has begun. The formula for calculating the response delay time is: in, This indicates the response delay time, in minutes. This indicates the moment when the leakage current begins to increase significantly. This indicates the moment when the relative humidity of the environment begins to rise.

[0047] In this embodiment, the response peak value in step S300 is defined as the maximum value of the leakage current amplitude reached during the humidity saturation stage. The humidity saturation stage refers to the period during which the relative humidity of the environment remains high after reaching its peak. The leakage current sequence during this period is extracted from the dynamic response data, and its maximum value is taken as the response peak value. The calculation formula is as follows: in, This indicates the peak response, measured in milliamperes (mA). This represents the magnitude of the leakage current at time t; Indicates the moment when the relative humidity of the environment reaches its peak; This indicates the duration of the humidity saturation period; in this embodiment, it is taken as 30 minutes.

[0048] In this embodiment, the recovery time in step S300 is defined as the time interval between the moment when the ambient relative humidity begins to decrease and the moment when the leakage current amplitude returns to a normal level. The normal level is defined as the average leakage current value of the device under dry conditions before the triggering event. The formula for calculating the recovery time is: in, Indicates recovery time, in minutes; This indicates the moment when the leakage current returns to normal levels; This indicates the moment when the relative humidity of the environment begins to decrease.

[0049] In this embodiment, the system pre-stores reference response features in step S300. These reference response features are standard response parameters obtained through multiple natural humidity change cycles of observation and statistics when the equipment is in good working order. The reference response features include reference delay time, reference peak value, and reference recovery time. The response feature parameters extracted in step one are compared with the corresponding reference response features to calculate the feature deviation. The feature deviation is expressed in the form of relative deviation, and the calculation formula is as follows: ; ; ; in, Indicates the amount of time delay deviation; Indicates the reference delay time; Indicates the peak deviation; Indicates the baseline peak value; Indicates the amount of recovery time deviation; Indicates the baseline recovery time.

[0050] The above three deviations together constitute the characteristic deviation vector, denoted as: The system retrieves the contamination event record for the target insulated equipment from the maintenance file based on the equipment code. The contamination event record contains an identifier indicating whether a new, different-colored covering has been added to the equipment. This identifier is a Boolean value; it is true if a new, different-colored covering is identified in step S100, and false otherwise. To facilitate subsequent model processing, this identifier is quantified into a numerical form, denoted as F, with a value of: The feature bias vector V and the soiling event record identifier F are input into the attribution discrimination model, which outputs the attribution judgment result R. The attribution discrimination model employs a lightweight rule-based model, requiring no complex training process. The model incorporates the following discrimination rules: (1) When all components in V are less than or equal to the first threshold When F=1, R is determined to be a surface contamination-dominated defect. In this embodiment, the first threshold... Set to 30%; (2) When there is at least one component in V that is greater than the second threshold When F=0, R is determined to be an internal crack-dominated defect. In this embodiment, the second threshold... Set to 80%; (3) When there is at least one component in V that is greater than the second threshold When F=1, R is determined to be a composite defect; (4) When all components in V are less than or equal to the first threshold When F=0, R is determined to be a normal device and no special processing is required; fifth, when the components in V are between the first threshold... Second threshold During this period, regardless of the value of F, R is considered abnormal and requires further observation; it is recommended to maintain enhanced monitoring.

[0051] First threshold Second threshold It is based on statistical analysis of historical operation and maintenance data. Preferably, Set as the upper limit of the 95% confidence interval for the characteristic deviation of historical normal samples; The lower limit of the 5% confidence interval for the characteristic deviation of historical severe defect samples is set.

[0052] The above discrimination rule can be expressed in mathematical form as follows: in, This represents the maximum value in the feature bias vector V. The output R of the attribution discriminant model serves as the input for subsequent steps, guiding operational decisions.

[0053] In this embodiment of the application, in step S300, the attribution discrimination model receives the feature bias vector. Then, call the current weight parameter vector. The comprehensive score of the current abnormal signal is calculated using a weighted summation algorithm. The calculation formula is: in, These represent the contribution weights of time delay features, amplitude features, and recovery features in the decision logic, respectively; the model will calculate the... The result is compared with a preset grading threshold range to determine the final attribution judgment. .

[0054] In this embodiment of the application, step S400, which stores the attribution judgment result and dynamic response data into a digital twin file, includes: Retrieve the pre-stored digital twin file based on the equipment identification of the target insulation equipment; The attribution judgment results, dynamic response data and corresponding pollution event records are linked and stored in a digital twin archive to form a time-series equipment health event record; Update the device status identifier in the digital twin archive based on the attribution judgment results; Generate corresponding maintenance suggestions based on the device status identifier.

[0055] In this embodiment of the application, the attribution judgment result is determined in step S400. This is stored as a core attribute in the digital twin archive's event_records; based on the attribution judgment results... Mapping generates device health status identifiers : like If the surface is dirty or abnormal, it needs to be observed. focus on; like If it is a composite defect, then abnormal; When generating operation and maintenance suggestions, the system prioritizes reading the attribution judgment results. To match specific handling strategies.

[0056] In this embodiment, in step S400, the system retrieves the pre-stored digital twin file of the target insulation device from the device asset database of the backend server based on the device code of the device. The digital twin file is a data set constructed with devices as units, used to store various types of information throughout the entire life cycle of the device.

[0057] In this embodiment of the application, the digital twin archive in step S400 is organized in JSON format, and its top-level data structure includes the following fields: Equipment code, a string type, uniquely identifies an insulating device; Basic equipment information, including model, production date, and commissioning date; status: Device current status identifier, enumeration type; : An array of device health event records, storing historical events in chronological order; Operations and maintenance personnel are advised to record an array.

[0058] In this embodiment, in step S400, the system associates and combines relevant data from the current diagnostic process to form a complete device health event record. This record contains the following five fields: The event's unique identifier is automatically generated by the system according to the rule of device code + timestamp; The event occurrence time, i.e., the moment when the trigger signal is generated in step S200; Contamination event record: The contamination event record generated in step S100 includes the time of contamination discovery, device code, and geographic location coordinates; Dynamic response data: Dynamic response data segments solidified and stored in step S200, including leakage current sequence, partial discharge sequence and ambient humidity sequence; Attribution judgment result: The attribution judgment result output by step S300 is strictly limited to one of the following five states: surface contamination-dominated defects, internal crack-dominated defects, composite defects, and equipment normal or abnormal and to be observed.

[0059] After encapsulating the above five fields into a JSON object, the system appends this object to the digital twin file. In the array. Write according to... Sort the field values ​​to ensure The array is always sorted in ascending order of time, forming a time-series record of device health events.

[0060] In an optional implementation, the attribution discrimination model in step S400 can also be constructed by extracting image features and environmental parameters to build a labeled dataset, training a supervised learning classifier such as a support vector machine or random forest, and having the model automatically learn the nonlinear mapping relationship between features and pollution causes to output the discrimination result.

[0061] In an optional implementation, the attribution discrimination model in step S400 can also be constructed by building a multi-source monitoring dataset containing time series dimensions, and training deep time series networks such as LSTM or Transformer to automatically extract dynamic evolution features and time series dependencies in the pollution development process, thereby achieving end-to-end intelligent discrimination of pollution causes.

[0062] In this embodiment of the application, in step S400, the system updates the device status identifier in the digital twin file based on the attribution judgment result output in step S300.

[0063] In this embodiment of the application, the device status identifier in step S400 adopts an enumeration type, and the correspondence between the value and the attribution judgment result is as follows: When the attribution judgment result is that the equipment is normal, the equipment status indicator is set to normal; When the attribution judgment result is a surface contamination-dominated defect or an anomaly to be observed, the equipment status indicator is set to concern. When the attribution judgment result is a complex defect, the equipment status indicator is set to abnormal. When the attribution judgment result is that the defect is dominated by internal cracks, the equipment status is set to severe. The system modifies the value of the status field in the digital twin file to the corresponding status identifier mentioned above and records the time of the status change.

[0064] In this embodiment of the application, in step S500, the system automatically generates corresponding operation and maintenance suggestions based on the updated device status identifier. The rules for generating operation and maintenance suggestions are as follows: When the device status is marked as normal, the generated maintenance suggestion is: The device is in good condition, maintain routine monitoring; When the equipment status is marked as "concerned" and the attribution judgment result is a surface contamination-dominated defect, the maintenance suggestion is: Surface contamination, it is recommended to arrange for live water flushing or cleaning, and check the bird deterrent device above. When the device status is marked as "concerned" and the attribution judgment result is "abnormal and pending observation", the generated operation and maintenance suggestion is: There is a minor abnormality. It is recommended to maintain enhanced monitoring and reassess after the next humidity change cycle. When the equipment status is marked as abnormal, the generated maintenance suggestion is: composite defect, there may be internal cracks under the surface dirt, it is recommended to prioritize power outage maintenance and replace the insulator; When the equipment status is marked as critical, the generated maintenance recommendation is: "Internal crack-dominated defects pose a significant operational risk. It is recommended to arrange an immediate power outage for replacement."

[0065] In this embodiment of the application, in step S500, the system stores the generated operation and maintenance suggestions into a digital twin archive. The suggestion is also pushed to the operations and maintenance personnel's mobile terminal or the to-do list of the operations and maintenance management system.

[0066] In this embodiment of the application, step S500, which updates the attribution judgment weight parameters based on the search results, includes: Feature vectors are extracted from dynamic response data. The feature vectors consist of response delay time deviation, response peak deviation, and recovery time deviation. The similarity between the feature vector and the feature vector of multiple historical cases in the cloud case library is calculated to obtain similar historical cases whose similarity exceeds a preset similarity threshold. Obtain the historical attribution judgment results and historical maintenance conclusions corresponding to similar historical cases; Compare the historical attribution judgment results and historical maintenance conclusions with the current secondary generation attribution judgment results; Based on the comparison results, the weight parameters used for subsequent attribution judgment in the attribution discrimination model are updated.

[0067] In this embodiment of the application, the feature vector in step S500 is composed of the three feature deviations calculated in step S300, namely: in, This indicates the amount of time delay deviation. This indicates the peak deviation. This represents the recovery time deviation. The eigenvector V is used to characterize the electrical response characteristics of the current defect.

[0068] In this embodiment, in step S500, the system sends the feature vector V to the cloud case library and performs similarity calculation with the feature vectors of multiple historical cases stored in the case library. The cloud case library is a historical case database deployed on a cloud server, and each historical case record contains the following fields: case identifier, device type, feature vector, historical attribution judgment result, and historical maintenance conclusion.

[0069] In this embodiment of the application, cosine similarity is used as a similarity metric in step S400, for the current feature vector V and the feature vector of any historical case. The formula for calculating cosine similarity is: in, This represents the cosine similarity between the current feature vector and the feature vector of the i-th historical case, with a value range of [-1, 1]. The closer the value is to 1, the more similar the two vectors are. This represents the j-th component of the current eigenvector; Let represent the j-th component of the feature vector of the i-th historical case. The system iterates through all historical cases in the cloud case database and calculates the similarity between each historical case and the current feature vector. When this happens, the historical case is classified as a similar historical case; a preset similarity threshold is used. Set to 0.85.

[0070] In an optional implementation, the similarity retrieval and matching algorithm of the cloud case library in step S500 can also convert case features into numerical vectors, calculate the Euclidean distance between the sample to be tested and all historical case vectors in the library, select the K nearest neighbors, and determine the final matching case based on their label voting results.

[0071] In an optional implementation, the similarity retrieval and matching algorithm of the cloud case library in step S500 can also use a deep learning model to transform the case data into high-dimensional semantic embedding vectors and store them in a dedicated vector database. By calculating metrics such as the cosine similarity between the query vector and the vectors in the library, millisecond-level efficient approximate nearest neighbor retrieval based on semantic understanding can be achieved.

[0072] For each similar historical case, the system retrieves the corresponding historical attribution judgment result from the cloud case database. and historical preservation conclusions Historical attribution judgment This is the result output by the attribution discrimination model when cases are added to the database; historical maintenance conclusions. It is the actual defect type confirmed after on-site verification or repair of the case, that is, the verified standard answer.

[0073] In this embodiment of the application, the historical maintenance conclusion in step S500 The values ​​are the same as the attribution judgment results, including five types: surface dirt-dominated defects (simple bird droppings), internal crack-dominated defects, composite defects, normal equipment, and abnormal defects to be observed.

[0074] The system will use the currently generated attribution judgment result. Historical attribution judgment results of similar historical cases and conclusions on historical preservation A comparison will be performed. The comparison results will fall into three categories: (1) Completely identical: when = = When the result is 1, it means that the current attribution judgment is consistent with the judgment and actual conclusion of similar historical cases, indicating that the judgment of the current discrimination model is accurate; (2) The model misjudged but the history was correct: when ≠ =and = When this occurs, it indicates that the attribution judgment of similar historical cases is accurate, but the current discrimination model makes an error in judging similar cases; (3) Historical misjudgment: when ≠ When this occurs, it indicates that there is an error in the attribution judgment of similar historical cases, and the case has been marked as needing correction when it was entered into the database.

[0075] In this embodiment of the application, the attribution discrimination model in step S500 adopts a weighted decision rule, and the weights of the three components of the feature bias vector V in the model are respectively... , , ,satisfy Regarding the parameter loading strategy, if the system is in its initial commissioning phase and no historical parameters are stored, the initial values ​​will be loaded by default. If the system is in the iterative operation phase, the historical weight values ​​stored after the previous S500 update are directly called as the current base weights. The system dynamically adjusts the loaded weight parameters based on the comparison results between the current feature vector and the cloud case library.

[0076] In this embodiment of the application, the weight update rule in step S500 is as follows: (1) When the comparison results are completely consistent, all weight parameters remain unchanged; (2) When the comparison result shows that the model is misjudged but the history is correct, the system analyzes the components of the current feature vector VV and the feature vectors of historical cases. The difference. If a feature component of the current case deviates significantly from the corresponding component of historical cases, and this deviation leads to misjudgment, then the weight of that component should be appropriately reduced. The weight adjustment amount δ is set to 0.05. After adjustment, all weights need to be normalized to keep the sum of weights at 1; (3) When the comparison result is a historical misjudgment, the system marks the similar historical case as pending review and does not participate in this weight update. At the same time, the case information is pushed to the backend administrator for manual review and correction. After the weight update is complete, the system will update the weight parameters. , , The data is stored in the configuration file of the attribution discrimination model for subsequent attribution judgments. Simultaneously, the system records this retrieval and update process in the operation log of the cloud-based case library, forming a traceable self-learning record.

[0077] Example 3 is an embodiment of the present invention. This embodiment differs from the first embodiment in that it provides an evaluation system for the operating status of external insulation equipment of overhead distribution lines.

[0078] It should be noted that the technical solution of the distribution network overhead line external insulation equipment operation status assessment system is based on the same concept as the above-mentioned distribution network overhead line external insulation equipment operation status assessment method. For details not described in detail in the technical solution of the distribution network overhead line external insulation equipment operation status assessment system in this embodiment, please refer to the description of the above-mentioned distribution network overhead line external insulation equipment operation status assessment method.

[0079] This embodiment provides a system for assessing the operational status of external insulation equipment for overhead power distribution lines, comprising: A dirt visual recognition module is used to acquire a surface image of the target insulation equipment and identify whether there is a new cover of a different color. In response to the identification of a new cover of a different color, a dirt event record containing a timestamp is generated, and the target insulation equipment is set to enhanced monitoring mode. The dynamic data capture module is used to collect the leakage current and partial discharge intensity of the target insulation device in real time in response to the target insulation device being in enhanced monitoring mode; and to acquire the dynamic response data of the target insulation device in the first natural humidity change cycle after the pollution event is recorded when either the leakage current or the partial discharge intensity is detected to exceed a preset threshold. The anomaly attribution analysis module is used to compare the dynamic response data with the pre-stored baseline response features and integrate the filth feature information in the filth event record to generate an attribution judgment result for the current anomaly signal. The data archiving module is used to store the attribution judgment results and the corresponding dynamic response data into the digital twin file of the target insulation device; The evaluation and optimization module is used to extract feature vectors based on the digital twin archives and perform similarity retrieval with the cloud case library. Based on the retrieval results, the attribution judgment weight parameters are updated to obtain an evaluation report on the operating status of the external insulation equipment.

[0080] This embodiment also provides an electronic device applicable to a method for assessing the operational status of external insulation equipment of overhead distribution lines, including: The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement a method for evaluating the operational status of external insulation equipment of overhead power distribution lines, as proposed in the above embodiments.

[0081] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a method for evaluating the operational status of external insulation equipment of overhead distribution lines as proposed in the above embodiments.

[0082] The storage medium proposed in this embodiment and the method for evaluating the operating status of external insulation equipment of overhead distribution lines proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0083] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0084] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for evaluating the operational status of external insulation equipment of overhead distribution lines, characterized in that, include: Acquire a surface image of the target insulation device and identify whether there is a new cover of a different color. In response to the identification of a new cover of a different color, generate a dirt event record with a timestamp and set the target insulation device to enhanced monitoring mode. In response to the target insulation device being in enhanced monitoring mode, the leakage current and partial discharge intensity of the target insulation device are collected in real time; in response to the detection that either the leakage current or the partial discharge intensity exceeds a preset threshold, the dynamic response data of the target insulation device within the first natural humidity change cycle after the pollution event is recorded are obtained. The dynamic response data is compared with the pre-stored baseline response features, and the filth feature information in the filth event record is fused to generate an attribution judgment result for the current abnormal signal. The attribution judgment results and the corresponding dynamic response data are stored in the digital twin file of the target insulation device; Feature vectors are extracted from the digital twin archives and similarity searches are performed with the cloud case library. Attribution judgment weight parameters are updated based on the search results to obtain an assessment report on the operating status of external insulation equipment.

2. The method for evaluating the operational status of external insulation equipment of overhead distribution lines as described in claim 1, characterized in that: The identification of whether there is a new cover of a different color includes: Collect the current surface image of the target insulation device at the current time point according to a preset cycle, and obtain historical surface images from the same viewpoint; Image registration and illumination correction are performed on the current surface image and the historical surface image to generate aligned and standardized current image and standardized historical image; The standardized current image and the standardized historical image are subjected to differential processing to extract the difference region; Extract the geometric and color features of the difference region, and compare the geometric and color features with a preset dirt feature threshold range; In response to the geometric features and the color features meeting the dirt feature threshold range, the differential region is determined to be a newly added heterochromatic covering.

3. The method for evaluating the operational status of external insulation equipment of overhead distribution lines as described in claim 1 or 2, characterized in that: Setting the target insulation device to enhanced monitoring mode includes: In response to determining that the differential region is a newly added different colored covering, the current time point is recorded as the time of contamination discovery, and the device code and geographical coordinates of the target insulation device are obtained, and a contamination event record containing the time of contamination discovery, the device code and the geographical coordinates are generated; The pollution event is recorded in the operation and maintenance file of the target insulation equipment, and the monitoring status identifier of the target insulation equipment is changed from routine monitoring to enhanced monitoring. The enhanced monitoring mode includes increasing the data acquisition frequency of leakage current and partial discharge intensity of the target insulation equipment, and extending the validity period of the pollution event records.

4. The method for evaluating the operational status of external insulation equipment of overhead distribution lines as described in claim 3, characterized in that: The acquisition of dynamic response data of the target insulation device during the first natural humidity change cycle after the contamination event is recorded includes: When the target insulation equipment is in enhanced monitoring mode, the leakage current amplitude and partial discharge pulse frequency of the target insulation equipment are continuously collected, and the ambient relative humidity data are collected simultaneously. In response to the leakage current amplitude exceeding a preset current threshold, a dynamic response data capture process is triggered. In response to the partial discharge pulse frequency exceeding a preset frequency threshold, a dynamic response data capture process is triggered; Starting from the triggering moment, record the process data of the leakage current amplitude and the partial discharge pulse frequency changing with the ambient relative humidity during the first natural humidity change cycle after the pollution event is recorded.

5. The method for evaluating the operational status of external insulation equipment of overhead distribution lines as described in claim 4, characterized in that: The triggering dynamic response data capture process includes: The real-time collected leakage current amplitude is compared with the preset current threshold, and the partial discharge pulse frequency is compared with the preset frequency threshold. In response to any monitored value exceeding the corresponding threshold, a trigger signal containing the type of the out-of-limit parameter and the time of the out-of-limit is generated; In response to the trigger signal, leakage current data, partial discharge data and synchronous ambient relative humidity data within a preset time period before and after the trigger time are solidified and stored to generate dynamic response data fragments to be analyzed. Initiate natural humidity change cycle monitoring to identify the first natural humidity change cycle after the recorded pollution event.

6. The method for evaluating the operational status of external insulation equipment of overhead distribution lines as described in claim 5, characterized in that: The generation of the attribution judgment result for the current abnormal signal includes: Response feature parameters are extracted from the dynamic response data, including response delay time, response peak value, and recovery time. The response feature parameters are compared with the pre-stored benchmark response features to obtain the feature deviation vector; Obtain the contamination event records of the target insulation equipment, and extract the identification information of whether there are any newly added different colored coverings in the contamination event records; The feature deviation vector and the identification information are input into the attribution discrimination model, and the attribution discrimination model outputs the attribution judgment result based on the magnitude of the feature deviation vector and the value of the identification information.

7. The method for evaluating the operational status of external insulation equipment of overhead distribution lines as described in claim 6, characterized in that: The step of updating the attribution judgment weight parameters based on the search results includes: Feature vectors are extracted from the dynamic response data. The feature vectors consist of response delay time deviation, response peak deviation, and recovery time deviation. The similarity between the feature vector and the feature vectors of multiple historical cases in the cloud case library is calculated to obtain similar historical cases whose similarity exceeds a preset similarity threshold. Obtain the historical attribution judgment results and historical maintenance conclusions corresponding to the similar historical cases; The historical attribution judgment results and the historical maintenance conclusions are compared with the currently generated attribution judgment results; Based on the comparison results, the weight parameters used for subsequent attribution judgment in the attribution discrimination model are updated.

8. A system for evaluating the operational status of external insulation equipment of overhead distribution lines, using the method described in any one of claims 1-7, characterized in that, include: A dirt visual recognition module is used to acquire a surface image of the target insulation equipment and identify whether there is a new cover of a different color. In response to the identification of a new cover of a different color, a dirt event record containing a timestamp is generated, and the target insulation equipment is set to enhanced monitoring mode. The dynamic data capture module is used to collect the leakage current and partial discharge intensity of the target insulation device in real time in response to the target insulation device being in enhanced monitoring mode; and to acquire the dynamic response data of the target insulation device in the first natural humidity change cycle after the pollution event is recorded when either the leakage current or the partial discharge intensity is detected to exceed a preset threshold. The anomaly attribution analysis module is used to compare the dynamic response data with the pre-stored baseline response features and integrate the filth feature information in the filth event record to generate an attribution judgment result for the current anomaly signal. The data archiving module is used to store the attribution judgment results and the corresponding dynamic response data into the digital twin file of the target insulation device; The evaluation and optimization module is used to extract feature vectors based on the digital twin archives and perform similarity retrieval with the cloud case library. Based on the retrieval results, the attribution judgment weight parameters are updated to obtain an evaluation report on the operating status of the external insulation equipment.

9. An electronic device, characterized in that, include: Memory, used to store programs; A processor for loading the program to perform the steps of the method as claimed in any one of claims 1-7.

10. A computer-readable storage medium storing a program, characterized in that, When the program is executed by a processor, it implements the steps of the method as described in any one of claims 1-7.

Images