A diagnostic method, system and storage medium for deteriorated insulators

Abstract

The present invention provides a method, system, and storage medium for diagnosing degraded insulators. This system performs dual diagnosis of the insulator under test using both electric field distribution curves and infrared images, ensuring accurate identification of degraded insulators. Furthermore, the system uses infrared images to determine and classify the degree of deterioration of the insulators. The system collects image and spectral data of the degraded insulators and classifies the degree of contamination based on a constructed contamination diagnosis model. Finally, the system uses a constructed knowledge graph model to analyze each labeled degraded insulator to determine the cause of deterioration and the corresponding solution. This system significantly improves the accuracy and efficiency of degraded insulator diagnosis, reduces the cost and labor intensity of manual diagnosis, ensures stable operation and power supply reliability of the power system, and is of great significance for preventing power accidents and reducing economic losses.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric power engineering and relates to a diagnostic method, system, equipment and storage medium for an insulator. Background Art

[0002] Insulators are common components on power transmission lines. Installed between conductors at different potentials or between conductors and grounded structures, they withstand voltage and mechanical stress and are exposed to harsh environments for extended periods. Insulators come in a wide variety of types and shapes, and while their structures and appearances vary significantly, they all consist of two main components: the insulating element and the connecting hardware. Degraded insulators are those whose overall insulation performance has deteriorated. This is due to long-term exposure to factors such as ultraviolet radiation, electrothermal stress, mechanical damage, chemical corrosion, environmental pollution, and temperature fluctuations. These factors can cause irreversible changes in the insulation material's properties, leading to a decrease in overall insulation performance, and even the appearance of zero-value insulators, increasing the risk of flashover and failure on transmission lines. Traditional diagnostic methods for degraded insulators rely on manual climbing or ground-based diagnosis, which is not only inefficient but also poses safety risks. Furthermore, they lack the ability to visually assess the extent of deterioration or the cause of deterioration. Consequently, diagnostic results are inaccurate, making it difficult for power operations and maintenance personnel to formulate maintenance plans. Summary of the Invention

[0003] In order to solve the problems of low diagnostic efficiency of deteriorated insulators, inability to intuitively judge the degree of deterioration and the cause of deterioration, and low accuracy of diagnostic results described in the background art, the present invention provides a diagnostic method, system and storage medium for deteriorated insulators.

[0004] The method of the present invention comprises:

[0005] Conduct live deterioration diagnosis on the insulators to be tested, and identify the deteriorated insulators in the insulator string based on the electric field distribution curve obtained from the diagnosis;

[0006] Acquire an infrared image of the insulator to be tested, compare the infrared image with an infrared image of a standard insulator, and identify the deteriorated insulator;

[0007] When the deteriorated insulator identified based on the electric field distribution curve and the deteriorated insulator identified based on the infrared image are the same deteriorated insulator, the deteriorated insulator is marked;

[0008] comparing the infrared image of the marked deteriorated insulator with the infrared image of a standard insulator to determine the degree of deterioration of the marked deteriorated insulator;

[0009] The images and spectral data of insulators with different pollution levels are input into the convolutional neural network model for training to build a pollution diagnosis model;

[0010] Using the image and spectral data of the marked deteriorated insulator, the contamination degree of the marked deteriorated insulator is determined according to the contamination diagnosis model;

[0011] A knowledge graph model is constructed based on the degree of deterioration of the deteriorated insulator, the degree of contamination of the deteriorated insulator, the inventory data of the deteriorated insulator, the regional environmental data of the deteriorated insulator, all deterioration causes of the deteriorated insulator, and the solutions corresponding to each deterioration cause.

[0012] Each marked deteriorated insulator is analyzed based on the knowledge graph model to obtain the degradation cause of the deteriorated insulator and the corresponding solution.

[0013] Furthermore, the method for diagnosing live degradation of the insulator to be tested is as follows: an electric field sensor is carried by a drone and flies to the outside of the shed edge of the insulator to be tested, and live degradation diagnosis from high potential to low potential is performed on the insulator to be tested, an electric field distribution curve is obtained, and the location of the degraded insulator is determined by the area where the electric field is distorted in the electric field distribution curve.

[0014] Furthermore, the method for obtaining the infrared image of the insulator to be tested is as follows: an infrared thermal imager is carried by a drone and photographs the insulator to be tested to obtain an infrared image, and the degraded insulator is identified based on the difference in surface temperature between the photographed infrared image and the infrared image of a normal insulator; the method for determining the degree of degradation of the marked degraded insulator is as follows: the temperature of the infrared image of the marked degraded insulator is compared with the temperature of the infrared image of the normal insulator; when the temperature of the marked degraded insulator is higher than the temperature of the normal insulator, the marked degraded insulator is a low-value insulator; when the temperature of the marked degraded insulator is lower than the temperature of the normal insulator, the marked degraded insulator is a zero-value insulator.

[0015] Furthermore, the construction process of the contamination diagnosis model includes:

[0016] Collect insulators of different pollution levels on transmission lines and acquire images and spectral data of the insulators using a hyperspectral imager;

[0017] Perform image preprocessing on the image and spectral data, including denoising, black and white correction, and multivariate scattering correction, to obtain preprocessed data;

[0018] The pre-processed data were analyzed using segmented principal component analysis to extract pollution characteristics;

[0019] The convolutional neural network mathematical model is trained and verified by combining the pollution characteristics and insulator data of different pollution levels to identify insulators of different pollution levels.

[0020] Furthermore, the calculation formula for the black and white correction is:

[0021]

[0022] Where R corrected Represents the corrected reflectivity, R raw Represents the original reflectivity, R max Represents the maximum value of the pixel value, that is, the average reflectance of the black calibration hyperspectral image, R min Represents the minimum value of the pixel value, that is, the average reflectance of the all-white calibration hyperspectral image;

[0023] The specific steps of the multivariate scatter correction are as follows:

[0024] Assume that the average value of all spectral data is an "ideal spectrum":

[0025]

[0026] Where h i is the spectral vector of a single sample, n is the number of selected sample spectral lines, and m is the number of spectral bands; The average spectrum vector obtained by averaging the spectral data of all samples at each band is also called the "ideal spectrum";

[0027] Then the spectral data of each group h i By performing a linear regression calculation with the ideal spectrum, the relative offset a of each spectrum can be obtained. i and baseline shift b i :

[0028]

[0029] Finally, after obtaining the baseline shift and offset, the baseline shift is subtracted from each set of sample spectral data and divided by the offset to correct the relative baseline tilt of the spectral data. Thus, the baseline shift and offset of each spectrum are corrected with reference to the standard spectrum.

[0030]

[0031] Where h i(MSC) is the spectral vector of a single sample after multivariate scattering correction.

[0032] Furthermore, the method for determining the contamination degree of the marked deteriorated insulator is as follows: a hyperspectral imager is carried out by a drone to perform aerial scanning of the deteriorated insulator to collect images and spectral data;

[0033] The image and spectrum data are input into a pollution diagnosis model to diagnose the pollution degree of the marked deteriorated insulators, and the deteriorated insulators are classified into insulators with pollution and insulators without pollution according to the pollution degree.

[0034] Furthermore, the construction process of the knowledge graph model includes:

[0035] Collect the inventory data of deteriorated insulators to obtain the installation time, production date, and service life of the deteriorated insulators, and use named entity recognition technology to construct a deteriorated insulator map;

[0036] Obtain the regional environmental data of the deteriorated insulators, obtain the environmental climate and air humidity of the deteriorated insulators, and use named entity recognition technology to construct a regional environmental map;

[0037] Based on the deterioration degree and contamination degree of the deteriorated insulators, combined with the deteriorated insulator atlas, the time extraction algorithm is used to identify the time information of the deteriorated insulators, obtain the service time and degradation trend of the deteriorated insulators, and construct a time series atlas;

[0038] Based on the deterioration degree and pollution degree of the deteriorated insulators, a degradation cause map is constructed by combining the deteriorated insulator map, regional environment map, time series map and all degradation causes;

[0039] Build a solution map based on the solutions corresponding to each degradation cause;

[0040] The knowledge graph is obtained by performing graph fusion, graph completion and graph reasoning on the degraded insulator graph, regional environment graph, degradation cause graph and solution graph based on the event extraction algorithm and TranSE algorithm.

[0041] The present invention also proposes a diagnostic system for degraded insulators, including a degraded insulator electric field identification module, a degraded insulator infrared identification module, a degraded insulator marking module, a degradation degree determination module, a contamination diagnosis model construction module, a contamination degree determination module, a knowledge graph model construction module, and a degraded insulator analysis module.

[0042] The degraded insulator electric field identification module is used to perform live degradation diagnosis on the insulator to be tested, and identify the degraded insulators in the insulator string based on the electric field distribution curve obtained by diagnosis.

[0043] The degraded insulator infrared recognition module is used to obtain an infrared image of the insulator to be tested, compare the infrared image with the infrared image of a standard insulator, and identify the degraded insulator.

[0044] The degraded insulator marking module is used to mark the degraded insulator when the degraded insulator identified based on the electric field distribution curve and the degraded insulator identified based on the infrared image are the same degraded insulator.

[0045] The degradation degree determination module is used to compare the infrared image of the marked degraded insulator with the infrared image of the standard insulator to determine the degradation degree of the marked degraded insulator.

[0046] The pollution diagnosis model building module is used to input images and spectral data of insulators with different pollution levels into a convolutional neural network model for training to build a pollution diagnosis model.

[0047] The contamination degree determination module is used to determine the contamination degree of the marked deteriorated insulator according to the contamination diagnosis model using the image and spectrum data of the marked deteriorated insulator.

[0048] The knowledge graph model construction module is used to construct a knowledge graph model based on the degradation degree of the degraded insulator, the pollution degree of the degraded insulator, the ledger data of the degraded insulator, the regional environmental data of the degraded insulator, all degradation causes of the degraded insulator and the solutions corresponding to each degradation cause.

[0049] The degraded insulator analysis module is used to analyze each marked degraded insulator based on the knowledge graph model to obtain the deterioration cause of the degraded insulator and the corresponding solution.

[0050] The present invention further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for diagnosing deteriorated insulators as described above is implemented.

[0051] Compared with the existing technology, the present invention ensures the accuracy of identifying deteriorated insulators by performing dual diagnosis on the insulator under test using electric field distribution curves and infrared images. It also determines and classifies the degree of deterioration of deteriorated insulators using infrared images. It collects images and spectral data of deteriorated insulators and classifies the degree of contamination of deteriorated insulators based on a constructed contamination diagnosis model. Finally, it analyzes each marked deteriorated insulator using a constructed knowledge graph model to obtain the cause of deterioration and the corresponding solution. The present invention can quickly cover a wide range of insulator strings on transmission lines. By combining electric field, infrared images, and image and spectral data, it can quickly and timely discover and locate deteriorated insulators during transmission line operation and inspection. It can also intuitively determine the degree of deterioration and the cause of deterioration of deteriorated insulators, greatly improving the accuracy and efficiency of deteriorated insulator diagnosis, reducing the cost and labor intensity of manual diagnosis, ensuring the stable operation and power supply reliability of the power system, and having great significance for preventing power accidents and reducing economic losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 Flow chart of the method of the present invention.

[0053] Figure 2 This is a system architecture diagram of the present invention.

[0054] Figure 3 This is an example diagram of the knowledge graph model of the present invention. DETAILED DESCRIPTION

[0055] In order to make the technical problems, technical solutions and beneficial effects to be solved by this application more clearly understood, this application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0056] Example 1

[0057] A method for diagnosing deteriorated insulators, the flow chart is as follows Figure 1 As shown, the details are as follows.

[0058] First, the insulator to be tested is diagnosed for deterioration under power conditions using an electric field sensor, and the deteriorated insulators in the insulator string are identified based on the electric field distribution curve obtained from the diagnosis.

[0059] Specifically, the electric field sensor can be an electro-optical sensor based on the Pockels effect; the electric field sensor is carried by a drone and flies to the outside of the shed edge of the insulator to be tested, and performs live degradation diagnosis of the insulator to be tested from high potential to low potential, obtains the electric field distribution curve, and determines the location of the degraded insulator through the area where the electric field is distorted in the electric field distribution curve.

[0060] Then, the insulator to be tested is photographed by an infrared thermal imager to obtain an infrared image, and the obtained infrared image is compared with the infrared image of a normal insulator to identify the deteriorated insulator.

[0061] Specifically, an infrared thermal imager is mounted on a drone and takes an infrared image of the insulator under test. Degraded insulators are identified based on the difference in surface temperature between the captured infrared image and the infrared image of a healthy insulator.

[0062] Then, the degraded insulators identified by the electric field sensor are compared and verified with the degraded insulators identified by the infrared thermal imager. If the comparison and verification are inconsistent, the insulator string to be tested is re-identified by the electric field sensor and the infrared thermal imager. If the comparison and verification are consistent, the degraded insulator is marked.

[0063] Dual diagnosis of the insulators to be tested is performed using electric field sensors and infrared thermal imagers to ensure the accuracy of identifying deteriorated insulators.

[0064] Next, the infrared image of the marked deteriorated insulator is compared with the infrared image of the normal insulator to judge the degree of deterioration of the deteriorated insulator, and the deteriorated insulator is classified into low-value insulators and zero-value insulators according to the degree of deterioration.

[0065] Specifically, the temperature of the infrared image of the marked deteriorated insulator is compared with that of the infrared image of the normal insulator. When the temperature of the marked deteriorated insulator is higher than the temperature of the normal insulator, the marked deteriorated insulator is a low-value insulator. When the temperature of the marked deteriorated insulator is lower than the temperature of the normal insulator, the marked deteriorated insulator is a zero-value insulator.

[0066] Next, images and spectral data of insulators with different degrees of contamination are collected through a hyperspectral imager and input into a convolutional neural network mathematical model for training and verification. Insulators with different degrees of contamination are identified and a contamination diagnosis model is constructed.

[0067] Specifically, the construction process of the pollution diagnosis model includes:

[0068] Collect insulators of different pollution levels on transmission lines and acquire images and spectral data of the insulators using a hyperspectral imager;

[0069] Perform image preprocessing on the image and spectral data, including denoising, black and white correction, and multivariate scattering correction, to obtain preprocessed data;

[0070] The calculation formula for black and white correction is:

[0071]

[0072] Where Rcorrected Represents the corrected reflectivity, R raw Represents the original reflectivity, R max Represents the maximum value of the pixel value, that is, the average reflectance of the black calibration hyperspectral image, R min Represents the minimum value of the pixel value, that is, the average reflectance of the all-white calibration hyperspectral image;

[0073] The specific steps of multivariate scatter correction are as follows:

[0074] Assume that the average value of all spectral data is an "ideal spectrum":

[0075]

[0076] Where h i is the spectral vector of a single sample, n is the number of selected sample spectral lines, and m is the number of spectral bands; The average spectrum vector obtained by averaging the spectral data of all samples at each band is also called the "ideal spectrum";

[0077] Then the spectral data of each group h i By performing a linear regression calculation with the ideal spectrum, the relative offset a of each spectrum can be obtained. i and baseline shift b i :

[0078]

[0079] Finally, after obtaining the baseline shift and offset, the baseline shift is subtracted from each set of sample spectral data and divided by the offset to correct the relative baseline tilt of the spectral data. Thus, the baseline shift and offset of each spectrum are corrected with reference to the standard spectrum.

[0080]

[0081] Where h i(MSC) is the spectral vector of a single sample after multivariate scattering correction;

[0082] The pre-processed data were analyzed using segmented principal component analysis to extract pollution characteristics;

[0083] The convolutional neural network mathematical model is trained and verified by combining the pollution characteristics and insulator data of different pollution levels to identify insulators of different pollution levels.

[0084] Next, the images and spectral data of the deteriorated insulators are collected using a hyperspectral imager, and the contamination degree of the deteriorated insulators is diagnosed based on the contamination diagnosis model. The deteriorated insulators are classified into insulators with contamination and insulators without contamination according to the contamination degree.

[0085] Specifically, the hyperspectral imager is carried by a drone to perform aerial scanning of deteriorated insulators and collect images and spectral data.

[0086] Next, based on the degree of degradation of the degraded insulators, the degree of contamination of the degraded insulators, the inventory data of the degraded insulators, the regional environmental data of the degraded insulators, all degradation causes of the degraded insulators, and the solutions corresponding to each degradation cause, the degradation causes of the degraded insulators are identified, and the corresponding solutions are output to construct a knowledge graph model.

[0087] Specifically, the construction process of the knowledge graph model includes:

[0088] Collect the inventory data of deteriorated insulators to obtain the installation time, production date, and service life of the deteriorated insulators, and use named entity recognition technology to construct a deteriorated insulator map;

[0089] Obtain the regional environmental data of the deteriorated insulators, obtain the environmental climate and air humidity of the deteriorated insulators, and use named entity recognition technology to construct a regional environmental map;

[0090] Based on the deterioration degree and contamination degree of the deteriorated insulators, combined with the deteriorated insulator atlas, the time extraction algorithm is used to identify the time information of the deteriorated insulators, obtain the service time and degradation trend of the deteriorated insulators, and construct a time series atlas;

[0091] Based on the deterioration degree and pollution degree of the deteriorated insulators, a degradation cause map is constructed by combining the deteriorated insulator map, regional environment map, time series map and all degradation causes;

[0092] Build a solution map based on the solutions corresponding to each degradation cause;

[0093] The knowledge graph is obtained by performing graph fusion, graph completion and graph reasoning on the degraded insulator graph, regional environment graph, degradation cause graph and solution graph based on the event extraction algorithm and TranSE algorithm.

[0094] An example diagram of the knowledge graph model is as follows Figure 3 shown.

[0095] Finally, each marked deteriorated insulator is analyzed based on the knowledge graph model to obtain the degradation cause of the deteriorated insulator and the corresponding solution.

[0096] Example 2

[0097] A diagnostic system for deteriorated insulators, the structure of which is shown in the following figure Figure 2As shown, it consists of a degraded insulator electric field recognition module, a degraded insulator infrared recognition module, a degraded insulator marking module, a degraded degree determination module, a contamination diagnosis model construction module, a contamination degree determination module, a knowledge graph model construction module, and a degraded insulator analysis module.

[0098] The degraded insulator electric field identification module uses electric field sensors to perform live degradation diagnosis on the insulator to be tested, and identifies the degraded insulators in the insulator string based on the electric field distribution curve obtained by diagnosis.

[0099] Specifically, the electric field sensor can be an electro-optical sensor based on the Pockels effect; the electric field sensor is carried by a drone and flies to the outside of the shed edge of the insulator to be tested, and performs live degradation diagnosis of the insulator to be tested from high potential to low potential, obtains the electric field distribution curve, and determines the location of the degraded insulator through the area where the electric field is distorted in the electric field distribution curve.

[0100] The deteriorated insulator infrared recognition module uses an infrared thermal imager to capture the insulator to be tested to obtain an infrared image, and compares the captured infrared image with the infrared image of a normal insulator to identify the deteriorated insulator.

[0101] Specifically, an infrared thermal imager is mounted on a drone and takes an infrared image of the insulator under test. Degraded insulators are identified based on the difference in surface temperature between the captured infrared image and the infrared image of a healthy insulator.

[0102] The deteriorated insulator marking module compares and verifies the deteriorated insulators identified by the electric field sensor with those identified by the infrared thermal imager. If the comparison and verification are inconsistent, the insulator string to be tested is re-identified by the electric field sensor and the infrared thermal imager. If the comparison and verification are consistent, the deteriorated insulator is marked.

[0103] Dual diagnosis of the insulators to be tested is performed using electric field sensors and infrared thermal imagers to ensure the accuracy of identifying deteriorated insulators.

[0104] The degradation degree determination module compares the infrared image of the marked deteriorated insulator with the infrared image of the normal insulator, judges the degradation degree of the deteriorated insulator, and classifies the deteriorated insulator into low-value insulators and zero-value insulators according to the degradation degree.

[0105] Specifically, the temperature of the infrared image of the marked deteriorated insulator is compared with that of the infrared image of the normal insulator. When the temperature of the marked deteriorated insulator is higher than the temperature of the normal insulator, the marked deteriorated insulator is a low-value insulator. When the temperature of the marked deteriorated insulator is lower than the temperature of the normal insulator, the marked deteriorated insulator is a zero-value insulator.

[0106] The pollution diagnosis model construction module uses a hyperspectral imager to collect images and spectral data of insulators with different pollution levels, and inputs them into the convolutional neural network mathematical model for training and verification. It identifies insulators with different pollution levels and constructs a pollution diagnosis model.

[0107] Specifically, the construction process of the pollution diagnosis model includes:

[0108] Collect insulators of different pollution levels on transmission lines and acquire images and spectral data of the insulators using a hyperspectral imager;

[0109] Perform image preprocessing on the image and spectral data, including denoising, black and white correction, and multivariate scattering correction, to obtain preprocessed data;

[0110] The calculation formula for black and white correction is:

[0111]

[0112] Where R corrected Represents the corrected reflectivity, R raw Represents the original reflectivity, R max Represents the maximum value of the pixel value, that is, the average reflectance of the black calibration hyperspectral image, R min Represents the minimum value of the pixel value, that is, the average reflectance of the all-white calibration hyperspectral image;

[0113] The specific steps of multivariate scatter correction are as follows:

[0114] Assume that the average value of all spectral data is an "ideal spectrum":

[0115]

[0116] Where h i is the spectral vector of a single sample, n is the number of selected sample spectral lines, and m is the number of spectral bands; The average spectrum vector obtained by averaging the spectral data of all samples at each band is also called the "ideal spectrum";

[0117] Then the spectral data of each group h i By performing a linear regression calculation with the ideal spectrum, the relative offset a of each spectrum can be obtained. i and baseline shift b i :

[0118]

[0119] Finally, after obtaining the baseline shift and offset, the baseline shift is subtracted from each set of sample spectral data and divided by the offset to correct the relative baseline tilt of the spectral data. Thus, the baseline shift and offset of each spectrum are corrected with reference to the standard spectrum.

[0120]

[0121] Where h i(MSC) is the spectral vector of a single sample after multivariate scattering correction;

[0122] The pre-processed data were analyzed using segmented principal component analysis to extract pollution characteristics;

[0123] The convolutional neural network mathematical model is trained and verified by combining the pollution characteristics and insulator data of different pollution levels to identify insulators of different pollution levels.

[0124] The pollution degree determination module collects images and spectral data of degraded insulators through a hyperspectral imager, diagnoses the pollution degree of the degraded insulators based on the pollution diagnosis model, and classifies the degraded insulators into insulators with pollution and insulators without pollution according to the pollution degree.

[0125] Specifically, the hyperspectral imager is carried by a drone to perform aerial scanning of deteriorated insulators and collect images and spectral data.

[0126] The knowledge graph model construction module identifies the degradation causes of deteriorated insulators based on the degradation degree of deteriorated insulators, the pollution degree of deteriorated insulators, the ledger data of deteriorated insulators, the regional environmental data of deteriorated insulators, all degradation causes of deteriorated insulators and the solutions corresponding to each degradation cause, and outputs the corresponding solutions to build a knowledge graph model.

[0127] Specifically, the construction process of the knowledge graph model includes:

[0128] Collect the inventory data of deteriorated insulators to obtain the installation time, production date, and service life of the deteriorated insulators, and use named entity recognition technology to construct a deteriorated insulator map;

[0129] Obtain the regional environmental data of the deteriorated insulators, obtain the environmental climate and air humidity of the deteriorated insulators, and use named entity recognition technology to construct a regional environmental map;

[0130] Based on the deterioration degree and contamination degree of the deteriorated insulators, combined with the deteriorated insulator atlas, the time extraction algorithm is used to identify the time information of the deteriorated insulators, obtain the service time and degradation trend of the deteriorated insulators, and construct a time series atlas;

[0131] Based on the deterioration degree and pollution degree of the deteriorated insulators, a degradation cause map is constructed by combining the deteriorated insulator map, regional environment map, time series map and all degradation causes;

[0132] Build a solution map based on the solutions corresponding to each degradation cause;

[0133] The knowledge graph is obtained by performing graph fusion, graph completion and graph reasoning on the degraded insulator graph, regional environment graph, degradation cause graph and solution graph based on the event extraction algorithm and TranSE algorithm.

[0134] An example diagram of the knowledge graph model is as follows Figure 3 shown.

[0135] The degraded insulator analysis module analyzes each marked degraded insulator based on the knowledge graph model to obtain the deterioration cause of the degraded insulator and the corresponding solution.

[0136] Example 3

[0137] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements a method for diagnosing a degraded insulator as described in Example 1 and a system for diagnosing a degraded insulator as described in Example 2.

[0138] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiment of the present application can be implemented in various computer languages, for example, object-oriented programming languages ​​Java, C++, Python, and literal translation scripting language JavaScript, etc.

[0139] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0140] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0142] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0143] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A method for diagnosing deteriorated insulators, characterized in that: include: Conduct live deterioration diagnosis on the insulators to be tested, and identify the deteriorated insulators in the insulator string based on the electric field distribution curve obtained from the diagnosis; Acquire an infrared image of the insulator to be tested, compare the infrared image with an infrared image of a standard insulator, and identify the deteriorated insulator; When the deteriorated insulator identified based on the electric field distribution curve and the deteriorated insulator identified based on the infrared image are the same deteriorated insulator, the deteriorated insulator is marked; comparing the infrared image of the marked deteriorated insulator with the infrared image of a standard insulator to determine the degree of deterioration of the marked deteriorated insulator; The images and spectral data of insulators with different pollution levels are input into the convolutional neural network model for training to build a pollution diagnosis model; The image and spectral data of the marked deteriorated insulator are used to determine the contamination degree of the marked deteriorated insulator according to a contamination diagnosis model. The method for determining the contamination degree of the marked deteriorated insulator comprises: using a hyperspectral imager carried by an unmanned aerial vehicle to perform aerial scanning of the deteriorated insulator to collect image and spectral data, inputting the image and spectral data into the contamination diagnosis model, and diagnosing the contamination degree of the marked deteriorated insulator. A knowledge graph model is constructed based on the degree of deterioration of the deteriorated insulator, the degree of contamination of the deteriorated insulator, the inventory data of the deteriorated insulator, the regional environmental data of the deteriorated insulator, all deterioration causes of the deteriorated insulator, and the solutions corresponding to each deterioration cause. Each marked deteriorated insulator is analyzed based on the knowledge graph model to obtain the degradation cause and corresponding solution of each marked deteriorated insulator.

2. The method for diagnosing deteriorated insulators according to claim 1, wherein: The method for diagnosing live degradation of an insulator to be tested comprises: an electric field sensor is carried by a drone and flies to the outside of the shed edge of the insulator to be tested, performing live degradation diagnosis from high potential to low potential on the insulator to be tested, obtaining an electric field distribution curve, and determining the location of the degraded insulator based on the area where the electric field is distorted in the electric field distribution curve.

3. The method for diagnosing deteriorated insulators according to claim 2, wherein: The infrared image acquisition method of the insulator to be tested comprises: an infrared thermal imager is carried by a drone and photographs the insulator to be tested to obtain an infrared image, and a deteriorated insulator is identified based on the difference in surface temperature between the infrared image obtained and the infrared image of a normal insulator; The method for determining the degree of degradation of the marked deteriorated insulator is as follows: comparing the temperature of the infrared image of the marked deteriorated insulator with that of the infrared image of a normal insulator; when the temperature of the marked deteriorated insulator is higher than the temperature of the normal insulator, the marked deteriorated insulator is a low-value insulator; and when the temperature of the marked deteriorated insulator is lower than the temperature of the normal insulator, the marked deteriorated insulator is a zero-value insulator.

4. The method for diagnosing deteriorated insulators according to claim 3, wherein: The construction process of the pollution diagnosis model includes: Collect insulators of different pollution levels on transmission lines and acquire images and spectral data of the insulators using a hyperspectral imager; Perform image preprocessing on the image and spectral data, including denoising, black and white correction, and multivariate scattering correction, to obtain preprocessed data; The pre-processed data were analyzed using segmented principal component analysis to extract pollution characteristics; The convolutional neural network mathematical model is trained and verified by combining the pollution characteristics and the data of insulators with different pollution levels to identify insulators with different pollution levels. The calculation formula for the black and white correction is: ， Where, represents the corrected reflectivity, represents the original reflectivity, Indicates the maximum value of the pixel value, that is, the average reflectance of the all-black calibration hyperspectral image. Represents the minimum value of the pixel value, that is, the average reflectance of the all-white calibration hyperspectral image; The specific steps of the multivariate scatter correction are as follows: Assume that the average value of all spectral data is an "ideal spectrum": ， Where, is the spectral vector of a single sample, n is the number of selected sample spectral lines, and m is the number of spectral bands; The average spectrum vector obtained by averaging the spectral data of all samples at each band, also known as the "ideal spectrum"; Then the spectral data of each group The relative offset of each spectrum can be obtained by performing a linear regression calculation with the ideal spectrum. and baseline shift : ； Finally, after obtaining the baseline shift and offset, the baseline shift is subtracted from each set of sample spectral data and divided by the offset to correct the relative baseline tilt of the spectral data. Thus, the baseline shift and offset of each spectrum are corrected with reference to the standard spectrum. ， Where, is the spectral vector of a single sample after multivariate scattering correction.

5. The method for diagnosing deteriorated insulators according to claim 4, characterized in that: Deteriorated insulators are classified into insulators with pollution degradation and insulators without pollution degradation according to the degree of pollution.

6. The method for diagnosing deteriorated insulators according to claim 5, characterized in that: The construction process of the knowledge graph model includes: Collect the inventory data of deteriorated insulators to obtain the installation time, production date, and service life of the deteriorated insulators, and use named entity recognition technology to construct a deteriorated insulator map; Obtain the regional environmental data of the deteriorated insulators, obtain the environmental climate and air humidity of the deteriorated insulators, and use named entity recognition technology to construct a regional environmental map; Based on the deterioration degree and contamination degree of the deteriorated insulators, combined with the deteriorated insulator atlas, the time extraction algorithm is used to identify the time information of the deteriorated insulators, obtain the service time and degradation trend of the deteriorated insulators, and construct a time series atlas; Based on the deterioration degree and pollution degree of the deteriorated insulators, a degradation cause map is constructed by combining the deteriorated insulator map, regional environment map, time series map and all degradation causes; Build a solution map based on the solutions corresponding to each degradation cause; The knowledge graph is obtained by performing graph fusion, graph completion and graph reasoning on the degraded insulator graph, regional environment graph, degradation cause graph and solution graph based on the event extraction algorithm and TranSE algorithm.

7. A diagnostic system for deteriorated insulators, characterized by: It includes a degraded insulator electric field identification module, a degraded insulator infrared identification module, a degraded insulator marking module, a degraded degree determination module, a contamination diagnosis model construction module, a contamination degree determination module, a knowledge graph model construction module, and a degraded insulator analysis module; The degraded insulator electric field identification module is used to perform live degradation diagnosis on the insulator to be tested and identify the degraded insulators in the insulator string based on the electric field distribution curve obtained by diagnosis; The degraded insulator infrared recognition module is used to obtain an infrared image of the insulator to be tested, compare the infrared image with the infrared image of a standard insulator, and identify the degraded insulator; The degraded insulator marking module is used to mark the degraded insulator when the degraded insulator identified based on the electric field distribution curve and the degraded insulator identified based on the infrared image are the same degraded insulator; The degradation degree determination module is configured to compare the infrared image of the marked deteriorated insulator with the infrared image of a standard insulator to determine the degradation degree of the marked deteriorated insulator; The pollution diagnosis model building module is used to input images and spectral data of insulators with different pollution levels into a convolutional neural network model for training to build a pollution diagnosis model; The contamination level determination module is configured to use the image and spectral data of the marked deteriorated insulator to determine the contamination level of the marked deteriorated insulator according to a contamination diagnosis model. The method for determining the contamination level of the marked deteriorated insulator is as follows: a hyperspectral imager is mounted on an unmanned aerial vehicle to perform an aerial scan of the deteriorated insulator to collect image and spectral data, and the image and spectral data are input into the contamination diagnosis model to diagnose the contamination level of the marked deteriorated insulator. The knowledge graph model construction module is used to construct a knowledge graph model based on the degree of degradation of the degraded insulator, the degree of contamination of the degraded insulator, the ledger data of the degraded insulator, the regional environmental data of the degraded insulator, all degradation causes of the degraded insulator, and the solutions corresponding to each degradation cause; The degraded insulator analysis module is used to analyze each marked degraded insulator based on the knowledge graph model to obtain the deterioration cause of the degraded insulator and the corresponding solution.

8. The deteriorated insulator diagnosis system according to claim 7, characterized in that: In the degraded insulator electric field identification module, the electric field sensor is carried by a drone and flies to the outside of the shed edge of the insulator to be tested, performs live degradation diagnosis of the insulator from high potential to low potential, obtains the electric field distribution curve, and determines the location of the degraded insulator based on the area where the electric field is distorted in the electric field distribution curve.

9. The deteriorated insulator diagnosis system according to claim 8, characterized in that: In the deteriorated insulator infrared recognition module, an infrared thermal imager is carried by a drone and photographs the insulator to be tested to obtain an infrared image. The deteriorated insulator is identified based on the difference in surface temperature between the infrared image obtained and the infrared image of a normal insulator. In the degradation degree determination module, the temperature of the infrared image of the marked deteriorated insulator is compared with the temperature of the infrared image of the normal insulator. When the temperature of the marked deteriorated insulator is higher than the temperature of the normal insulator, the marked deteriorated insulator is a low-value insulator. When the temperature of the marked deteriorated insulator is lower than the temperature of the normal insulator, the marked deteriorated insulator is a zero-value insulator.

10. The deteriorated insulator diagnosis system according to claim 9, characterized in that: In the contamination diagnosis model construction module, the construction process of the contamination diagnosis model includes: Collect insulators of different pollution levels on transmission lines and acquire images and spectral data of the insulators using a hyperspectral imager; Perform image preprocessing on the image and spectral data, including denoising, black and white correction, and multivariate scattering correction, to obtain preprocessed data; The pre-processed data were analyzed using segmented principal component analysis to extract pollution characteristics; The convolutional neural network mathematical model is trained and verified by combining the pollution characteristics and the data of insulators with different pollution levels to identify insulators with different pollution levels. The calculation formula for the black and white correction is: ， Where, represents the corrected reflectivity, represents the original reflectivity, Indicates the maximum value of the pixel value, that is, the average reflectance of the all-black calibration hyperspectral image. Represents the minimum value of the pixel value, that is, the average reflectance of the all-white calibration hyperspectral image; The specific steps of the multivariate scatter correction are as follows: Assume that the average value of all spectral data is an "ideal spectrum": ， Where, is the spectral vector of a single sample, n is the number of selected sample spectral lines, and m is the number of spectral bands; The average spectrum vector obtained by averaging the spectral data of all samples at each band, also known as the "ideal spectrum"; Then the spectral data of each group The relative offset of each spectrum can be obtained by performing a linear regression calculation with the ideal spectrum. and baseline shift : ； Finally, after obtaining the baseline shift and offset, the baseline shift is subtracted from each set of sample spectral data and divided by the offset to correct the relative baseline tilt of the spectral data. Thus, the baseline shift and offset of each spectrum are corrected with reference to the standard spectrum. ， Where, is the spectral vector of a single sample after multivariate scattering correction.

11. The deteriorated insulator diagnosis system according to claim 10, characterized in that: In the pollution degree determination module, a hyperspectral imager is carried out by a drone to perform aerial scanning of deteriorated insulators to collect images and spectral data; The image and spectrum data are input into a pollution diagnosis model to diagnose the pollution degree of the marked deteriorated insulators, and the deteriorated insulators are classified into insulators with pollution and insulators without pollution according to the pollution degree.

12. The deteriorated insulator diagnosis system according to claim 11, characterized in that: In the knowledge graph model construction module, the ledger data of deteriorated insulators is collected to obtain the installation time, production date, and service life of the deteriorated insulators, and the deteriorated insulator graph is constructed using named entity recognition technology; Obtain the regional environmental data of the deteriorated insulators, obtain the environmental climate and air humidity of the deteriorated insulators, and use named entity recognition technology to construct a regional environmental map; Based on the deterioration degree and contamination degree of the deteriorated insulators, combined with the deteriorated insulator atlas, the time extraction algorithm is used to identify the time information of the deteriorated insulators, obtain the service time and degradation trend of the deteriorated insulators, and construct a time series atlas; Based on the deterioration degree and pollution degree of the deteriorated insulators, a degradation cause map is constructed by combining the deteriorated insulator map, regional environment map, time series map and all degradation causes; Build a solution map based on the solutions corresponding to each degradation cause; The knowledge graph is obtained by performing graph fusion, graph completion and graph reasoning on the degraded insulator graph, regional environment graph, degradation cause graph and solution graph based on the event extraction algorithm and TranSE algorithm.

13. A computer-readable storage medium storing a computer program, wherein: When the computer program is executed by a processor, the method for diagnosing a deteriorated insulator according to any one of claims 1 to 6 is implemented.

Images