Transformer substation GIS equipment fault positioning image detection method and system

Abstract

The invention discloses a transformer substation GIS equipment fault positioning image detection method and system, and relates to the technical field of GIS equipment fault detection, and the method comprises the following steps: obtaining an infrared thermodynamic diagram of GIS equipment of a transformer substation; partitioning components of the GIS equipment in the infrared thermodynamic diagram to obtain a to-be-analyzed image; extracting infrared features of the GIS device based on the to-be-analyzed image; the temperature of the GIS equipment in the infrared thermodynamic diagram is restored through the environment temperature, and normal infrared features are learned based on the power load and the environment temperature; judging whether the GIS equipment has a thermal fault or not by combining the power load, the environment temperature and the infrared characteristics; the method is used for solving the problem that whether the GIS equipment has a fault or not cannot be effectively recognized due to the fact that a detection mode is too traditional or a detection model is poor in universality in an existing GIS equipment fault detection technology.

Description

Technical Field

[0001] The present invention relates to the technical field of GIS equipment fault detection, and specifically to a method and system for detecting and locating faults in GIS equipment of a substation by means of images. Background Art

[0002] The GIS equipment fault detection technology refers to a comprehensive technology for non-contact and automated fault identification and location of gas-insulated switchgear in a substation. Its core objective is to quickly detect potential or existing abnormal states of equipment through intelligent means and accurately determine the fault location, thereby improving the safety and operation and maintenance efficiency of the power system.

[0003] Traditional GIS equipment fault detection technologies usually detect faults in GIS equipment by means of manual visual inspection or power-off tests, or by presetting thresholds to monitor in real time whether a certain parameter in the GIS equipment exceeds the standard. The above methods are too simple, and the results of fault detection are prone to deviation. At the same time, existing GIS equipment fault detection technologies also detect whether there are faults in GIS equipment by detecting a specific component in the GIS equipment. Since GIS equipment is composed of multiple equipment components, if only a single component is detected separately, a detection model needs to be independently set for each component, and the versatility is poor. For example, in the patent application with the publication number CN114719822A, a "method for detecting faults in GIS equipment based on geometric methods" is disclosed. This solution evaluates whether there are faults in GIS equipment by independently analyzing the conductive rod in the GIS equipment, and no evaluation method is set for the remaining components. If there are faults in the remaining components, this solution cannot effectively identify them. Existing GIS equipment fault detection technologies also have problems such as overly traditional detection methods or poor versatility of detection models, resulting in the inability to effectively identify whether there are faults in GIS equipment. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems in the prior art to some extent. By obtaining the infrared thermal map of the GIS equipment in the substation, partitioning the components of the GIS equipment in the infrared thermal map to obtain the image to be analyzed, then extracting the infrared features of the GIS equipment based on the image to be analyzed. At the same time, based on normal historical monitoring data, reduction analysis is performed on Tmax and Tmin, and then the infrared features of the GIS equipment are re-analyzed after the reduction analysis of Tmax and Tmin. Then, learning is carried out on the normal infrared features based on the power load and ambient temperature. Finally, combining the power load, ambient temperature, and infrared features, it is determined whether there is a thermal fault in the GIS equipment, so as to solve the problems that existing GIS equipment fault detection technologies have overly traditional detection methods or poor versatility of detection models, resulting in the inability to effectively identify whether there are faults in GIS equipment.

[0005] To achieve the above object, in a first aspect, the present application provides a method for detecting images of faults in GIS equipment of a substation, including the following steps: Obtain an infrared thermal map of the GIS equipment of the substation; Partition the components of the GIS equipment in the infrared thermal map to obtain an image to be analyzed; Extract the infrared features of the GIS equipment based on the image to be analyzed; Obtain the power load and ambient temperature in normal historical monitoring data, restore the temperature of the GIS equipment in the infrared thermal map through the ambient temperature, and then learn the normal infrared features based on the power load and ambient temperature; Combine the power load, ambient temperature, and infrared features to determine whether there is a thermal fault in the GIS equipment.

[0006] Further, obtaining the infrared thermal map of the GIS equipment of the substation includes the following sub-steps: Perform real-time monitoring of the GIS equipment through an infrared thermal imager; The infrared thermal imager is placed in four directions: directly in front of, directly behind, directly to the left of, and directly to the right of the GIS equipment; Name the image captured by the infrared thermal imager as the infrared thermal map.

[0007] Further, partitioning the components of the GIS equipment in the infrared thermal map to obtain an image to be analyzed includes the following sub-steps: The GIS equipment is composed of a first number of components; Manually frame and partition the components in the infrared thermal map, and each framed partition is an image to be analyzed, and each image to be analyzed corresponds to a monitoring area; For manual partitioning, only the infrared thermal maps in four directions need to be partitioned for the first time. When partitioning subsequently, directly obtain the monitoring area corresponding to the first partitioning.

[0008] Further, extracting the infrared features of the GIS equipment based on the image to be analyzed includes the following sub-steps: For any image to be analyzed, convert the image to be analyzed into a grayscale image, named the grayscale image to be analyzed; Mark the pixel point at the nth row and mth column in the grayscale image to be analyzed as P(n, m), where both n and m are positive integers and (n, m) is the serial number of P; Obtain the corresponding grayscale value at P(n, m), represented by the symbol H(n, m), find the maximum and minimum values in H(n, m), and mark them as Hmax and Hmin respectively. Obtain the maximum and minimum values of the temperature in the image to be analyzed through the infrared thermal imager, and mark them as Tmax and Tmin respectively; Calculate the temperature of each P(n,m) through the formula where T(n,m) is the temperature of P(n,m); Analyze the infrared characteristics of the image to be analyzed based on T(n,m).

[0009] Furthermore, analyzing the infrared characteristics of the image to be analyzed based on T(n,m) includes the following sub-steps: Find the P(n,m) where H(n,m) = Hmax, mark it as the heat source point, count all the heat source points to obtain the heat source set; Sort and number T(n,m) in ascending order, represented by the symbol G i where i is a positive integer and i is the serial number of G; Take i as the X-axis and G i as the Y-axis to establish a plane rectangular coordinate system, named the infrared characteristic diagram, and input G i into the infrared characteristic diagram according to i; Conduct a linear regression analysis on the infrared characteristic diagram, obtain the slope of the linear regression function, named the characteristic slope, and the heat source set and the characteristic slope are the infrared characteristics.

[0010] Furthermore, obtain the power load and ambient temperature in the normal historical monitoring data, restore the temperature of the GIS equipment in the infrared thermal diagram through the ambient temperature, and then learn the normal infrared characteristics based on the power load and ambient temperature, including the following sub-steps: Conduct a restoration analysis on Tmax and Tmin based on the normal historical monitoring data; After conducting a restoration analysis on Tmax and Tmin, re-analyze the infrared characteristics of the GIS equipment, and then learn the normal infrared characteristics based on the power load and ambient temperature.

[0011] Furthermore, conducting a restoration analysis on Tmax and Tmin based on the normal historical monitoring data includes the following sub-steps: Analyze any monitoring area to obtain the normal historical monitoring data. The historical monitoring data records the power load, ambient temperature, Tmax, and Tmin at any time, marked as historical load, historical temperature, Lmax, and Lmin respectively. The historical monitoring data of each monitoring area is independent of each other; Find the historical load with the most repeated times of the historical load, and integrate the corresponding historical monitoring data into the restoration reference data group; Analyze the historical monitoring data in the restoration reference data group, number the historical temperature, represented by the symbol R j where j is a positive integer and j is the serial number of R. Mark the corresponding Lmax and Lmin of R j as EA respectivelyj and EB j ; Calculate R j -EA j and R j -EB j , get the maximum and minimum ambient temperature difference of each historical monitoring data, mark them as Δt1 and Δt2 respectively, find the average value of Lmax when Δt1 is zero, and find the average value of Lmin when Δt2 is zero, mark them as F1 and F2 respectively, and calculate EA j -F1, get the maximum temperature rise of each historical monitoring data, marked as St1, calculate EA j -F2, get the minimum temperature rise of each historical monitoring data, marked as St2; A plane rectangular coordinate system is established with Δt1 as the horizontal axis and St1 as the vertical axis, named as the maximum temperature influence diagram, and St1 is recorded in the maximum temperature influence diagram according to the corresponding Δt1; a plane rectangular coordinate system is established with Δt2 as the horizontal axis and St2 as the vertical axis, named as the minimum temperature influence diagram, and St2 is recorded in the minimum temperature influence diagram according to the corresponding Δt2; Perform polynomial regression analysis on the maximum temperature influence diagram and the minimum temperature influence diagram, and name the regression functions as the maximum temperature influence function and the minimum temperature influence function, respectively. This completes the analysis of the restored reference data set. For all historical monitoring data, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively, solve the maximum temperature influence value and the minimum temperature influence value, subtract Lmax from the maximum temperature influence value to obtain the restored value of Lmax, marked as Umax, subtract Lmin from the minimum temperature influence value to obtain the restored value of Lmin, marked as Umin.

[0012] Furthermore, after restoring and analyzing Tmax and Tmin, the infrared characteristics of the GIS equipment are re-analyzed, and then the normal infrared characteristics are learned based on the power load and ambient temperature, including the following sub-steps: After restoring Lmax and Lmin in the historical monitoring data, rename the historical monitoring data as restored monitoring data, use Umax in the restored monitoring data as Tmax, and Umin as Tmin, and re-analyze the infrared features of the image to be analyzed; The heat source sets of all restored monitoring data are merged to obtain the characteristic distribution set; A plane rectangular coordinate system is established with power load as the X-axis and characteristic slope as the Y-axis, named as characteristic distribution diagram, and the characteristic slope of the restored monitoring data is entered into the characteristic distribution diagram according to power load; Perform a linear regression analysis on the characteristic distribution diagram, name the regression function as the characteristic distribution function, and at the same time obtain the maximum value of the residual of the characteristic distribution function, which is marked as the fluctuation value; Each monitoring area corresponds to a fluctuation value, a characteristic distribution function, and a characteristic distribution set.

[0013] Furthermore, combining the power load, ambient temperature, and infrared characteristics, determining whether there is a thermal fault in the GIS device includes the following sub-steps: Real-time monitor the power load and ambient temperature of the substation, and mark them as the real-time load and real-time temperature respectively; For any monitoring area, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively, solve the maximum temperature influence value and the minimum temperature influence value, subtract the maximum temperature influence value from Lmax to obtain the reduced value of Lmax, which is marked as Vmax, and subtract the minimum temperature influence value from Lmin to obtain the reduced value of Lmin, which is marked as Vmin; Analyze the infrared characteristics of the monitoring area through Vmax and Vmin, including the heat source set and the characteristic slope, which are named the real-time set and the real-time slope respectively; Mark the real-time set and the characteristic distribution set as D1 and D2 respectively. If D1 ∈ D2, output a heat source normal signal, otherwise output a heat source abnormal signal; Substitute the real-time load into the characteristic distribution function, name the obtained value as the ideal characteristic, increase the ideal characteristic by the fluctuation value to obtain the maximum characteristic, and decrease the ideal characteristic by the fluctuation value to obtain the minimum characteristic; Judge whether the real-time slope is between the maximum characteristic and the minimum characteristic. If so, output a characteristic normal signal, otherwise output a characteristic abnormal signal; If a heat source abnormal signal or a characteristic abnormal signal is output, mark that there is a fault in the corresponding monitoring area.

[0014] In a second aspect, the present application provides a substation GIS device fault location image detection system, including an infrared monitoring module, an image partitioning module, a feature analysis module, a feature learning module, and a fault judgment module; the infrared monitoring module, the image partitioning module, the feature analysis module, and the fault judgment module are respectively connected to the feature learning module for data connection; The infrared monitoring module is used to obtain the infrared thermal map of the GIS device of the substation; The image partitioning module is used to partition the components of the GIS device in the infrared thermal map to obtain an image to be analyzed; The feature analysis module is used to extract the infrared features of the GIS device based on the image to be analyzed; The feature learning module is used to obtain the power load and ambient temperature in normal historical monitoring data, restore the temperature of GIS equipment in the infrared thermal map through the ambient temperature, and then learn the normal infrared features based on the power load and ambient temperature; The fault judgment module is used to judge whether there is a thermal fault in the GIS equipment by combining the power load, ambient temperature and infrared features.

[0015] Advantages of the present invention: By obtaining the infrared thermal map of GIS equipment in a substation, partitioning the components of the GIS equipment in the infrared thermal map to obtain an image to be analyzed, and then extracting the infrared features of the GIS equipment based on the image to be analyzed. At the same time, based on normal historical monitoring data, Tmax and Tmin are restored and analyzed. The advantage is that since the installation position of the infrared thermal imager is unchanged, the positions of different components in the GIS equipment in the infrared thermal map remain unchanged. Therefore, after the positions of the components are partitioned manually for the first time, several monitoring areas are obtained for fault monitoring of each component. During the self-heating process of the GIS equipment, it is also affected by the ambient temperature, which causes the temperature of the GIS equipment to change. It is necessary to eliminate the influence of the ambient temperature on the temperature of the GIS equipment to most accurately evaluate whether there is a fault in the GIS equipment, improving the accuracy and rationality of GIS equipment fault detection; After restoring and analyzing Tmax and Tmin, the present invention re-analyzes the infrared features of GIS equipment, then learns the normal infrared features based on the power load and ambient temperature, and finally combines the power load, ambient temperature and infrared features to judge whether there is a thermal fault in the GIS equipment. The advantage is that the infrared features of the GIS equipment in the infrared thermal map are extracted, including the heat source set and the characteristic slope. The heat source set represents the heat generation points of the GIS equipment, and the positions of the heat sources are fixed, so the heat generation points are usually unchanged. The characteristic slope reveals the temperature change range on the surface of the GIS equipment. If there is an abnormally high temperature, it will cause a large change in the characteristic slope, thereby judging whether there is a fault in the GIS equipment, improving the accuracy and effectiveness of GIS equipment fault detection. Description of the Drawings

[0016] Figure 1 is the principle block diagram of the system of the present invention; Figure 2 is the image to be analyzed of the present invention; Figure 3 is the infrared feature map of the present invention; Figure 4 is the maximum temperature influence map of the present invention; Figure 5 is the feature distribution map of the present invention; Figure 6This is the flowchart of the steps of the method of the present invention. Detailed implementation manners

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] Example 1, please refer to Figure 1 As shown, the present application provides a substation GIS equipment fault location image detection system, including an infrared monitoring module, an image partitioning module, a feature analysis module, a feature learning module, and a fault judgment module; the infrared monitoring module, the image partitioning module, the feature analysis module, and the fault judgment module are respectively connected to the feature learning module for data connection; The infrared monitoring module is used to obtain the infrared thermal map of the GIS equipment of the substation; The infrared monitoring module is configured with an infrared monitoring strategy, and the infrared monitoring strategy includes: Perform real-time monitoring of the GIS equipment through an infrared thermal imager; Place the infrared thermal imager in four directions: directly in front of, directly behind, directly to the left of, and directly to the right of the GIS equipment; Name the image captured by the infrared thermal imager as the infrared thermal map; In practical applications, placing the infrared thermal imager in four directions: directly in front of, directly behind, directly to the left of, and directly to the right of the GIS equipment is to ensure obtaining the thermal images of the GIS equipment in all directions. Under permitted conditions, infrared thermal imagers can also be installed for monitoring in the directly above and directly below directions, and only the pixels with heat are retained in the infrared thermal map, and the rest of the background is automatically processed.

[0019] The image partitioning module is used to partition the components of the GIS equipment in the infrared thermal map to obtain the image to be analyzed; The image partitioning module is configured with an image partitioning strategy, and the image partitioning strategy includes: The GIS equipment is composed of components with the number of the first component; Manually frame and partition the components in the infrared thermal map, and each framed partition is an image to be analyzed, and each image to be analyzed corresponds to a monitoring area; For manual partitioning, only the infrared thermal maps in four directions need to be partitioned for the first time. When partitioning later, directly obtain the monitoring area corresponding to the first partitioning; In practical applications, the number of the first components is determined by the GIS device. When the infrared thermal image is captured for the first time, the components of the GIS device in the infrared thermal image are manually framed and partitioned. Since the infrared thermal image is captured by a fixed camera position and has the same resolution, the boundary of the image to be analyzed after partitioning is the boundary of the monitoring area. When analyzing later, the components of the GIS device in the infrared thermal images captured from different orientations can be automatically partitioned directly based on the boundary of the monitoring area.

[0020] The feature analysis module is used to extract the infrared features of the GIS device based on the image to be analyzed; the feature analysis module includes a temperature calculation unit and a feature analysis unit; The temperature calculation unit is configured with a temperature calculation strategy, and the temperature calculation strategy includes: Please refer to Figure 2 As shown, for any image to be analyzed, convert the image to be analyzed into a grayscale image, named the grayscale image to be analyzed; In practical applications, this embodiment takes Figure 2 the image to be analyzed shown as an example to illustrate the fault detection process of the GIS device. The analysis process of each image to be analyzed is the same, only the final judgment basis obtained by analysis is different; Mark the pixel point at the nth row and the mth column in the grayscale image to be analyzed as P(n, m), where both n and m are positive integers and (n, m) is the serial number of P; Obtain the corresponding grayscale value at P(n, m), represented by the symbol H(n, m), find the maximum value and the minimum value in H(n, m), and mark them as Hmax and Hmin respectively. Obtain the maximum value and the minimum value of the temperature in the image to be analyzed through an infrared thermal imager, and mark them as Tmax and Tmin respectively; Through the formula calculate the temperature of each P(n, m), where T(n, m) is the temperature of P(n, m); In practical applications, when Hmax and Hmin are obtained as 252 and 30 respectively, select the corresponding image to be analyzed in the infrared thermal imager to obtain Tmax and Tmin of the image to be analyzed, which are 46.2°C and 22°C respectively. Taking P(158, 236) as an example, when H(158, 236) is obtained as 104, substitute it into the calculation to get T(158, 236) as 30.1°C. The calculation result is reserved to one decimal place, and calculate T(n, m) of each P(n, m); The feature analysis unit is used to analyze the infrared features of the image to be analyzed based on T(n, m); The feature analysis unit is configured with a feature analysis strategy, and the feature analysis strategy includes: Find P(n, m) where H(n, m)=Hmax, mark it as the heat source point, and count all the heat source points to obtain the heat source set; Sort and number T(n,m) in ascending order, denoted by the symbol G i where i is a positive integer and i is the serial number of G; Please refer to Figure 3 as shown. Take i as the X-axis and G i as the Y-axis to establish a plane rectangular coordinate system, named the infrared feature map, and input G i into the infrared feature map according to i; Conduct a linear regression analysis on the infrared feature map to obtain the slope of the linear regression function, named the feature slope. The heat source set and the feature slope are the infrared features; In practical applications, find P(n,m) where H(n,m) = Hmax = 252, mark it as the heat source point, and statistically obtain the heat source set, including 2590 pixel points such as P(186,330), P(187,330), P(188,330), P(189,330), P(188,331), and P(187,329); construct the infrared feature map as Figure 3 shown. Through linear regression analysis, the feature slope is obtained as 0.0004. In all coordinate systems in this embodiment, since there are too many data points to facilitate observation, only some coordinate points are shown in this embodiment.

[0021] The feature learning module is used to obtain the power load and environmental temperature in normal historical monitoring data, restore the temperature of GIS devices in the infrared thermal map through the environmental temperature, and then learn the normal infrared features based on the power load and environmental temperature; the feature learning module includes a restoration analysis unit and a feature learning unit; The restoration analysis unit is used to conduct a restoration analysis on Tmax and Tmin based on normal historical monitoring data; The restoration analysis unit is configured with a restoration analysis strategy, and the restoration analysis strategy includes: Analyze any monitoring area to obtain normal historical monitoring data. The historical monitoring data records the power load, environmental temperature, Tmax, and Tmin at any time, respectively marked as historical load, historical temperature, Lmax, and Lmin. The historical monitoring data of each monitoring area is independent of each other; Find the historical load with the most repeated times of historical load, and integrate the corresponding historical monitoring data into a restoration reference data group; Analyze the historical monitoring data in the restoration reference data group, number the historical temperature, denoted by the symbol R j where j is a positive integer and j is the serial number of R, and mark the corresponding Lmax and Lmin of R j as EA j and EB j respectively; In practical applications, obtain Figure 2 the historical detection data of the corresponding detection area. For example, there are 1,826 pieces of historical detection data with a historical load of 23 MW, which occupies the largest proportion. Then, integrate the historical detection data with a historical load of 23 MW into a restored reference data group, number the historical temperatures of the historical monitoring data in the restored reference data group to obtain R j , and at the same time mark to obtain EA j and EB j , 1 ≤ j ≤ 1,826; Calculate R j -EA j and R j -EB j , obtain the maximum ambient temperature difference and the minimum of each piece of historical monitoring data, and mark them as Δt1 and Δt2 respectively. Find the average value of Lmax when Δt1 is zero, and at the same time find the average value of Lmin when Δt2 is zero, and mark them as F1 and F2 respectively. Calculate EA j -F1, obtain the maximum temperature rise affected by each piece of historical monitoring data, and mark it as St1. Calculate EA j -F2, obtain the minimum temperature rise affected by each piece of historical monitoring data, and mark it as St2; In practical applications, the influence of the ambient temperature on the temperature of the GIS device depends on the difference between the ambient temperature and the temperature of the GIS device, rather than the ambient temperature itself. If the ambient temperature is greater than Lmax, it means that the ambient temperature will affect the GIS device and cause Lmax to rise. If the ambient temperature is less than Lmax, it means that the ambient temperature will affect the GIS device and cause Lmax to fall. The influence of the ambient temperature on Lmin is the same. Therefore, it is necessary to calculate the maximum ambient temperature difference Δt1 and the minimum ambient temperature difference Δt2. For example, in one piece of historical detection data, the historical load is 23 MW, the ambient temperature is 32 °C, Lmax is 35 °C, and Lmin is 24 °C. At this time, calculate Δt1 = 32 °C - 35 °C = -3 °C, Δt2 = 32 °C - 24 °C = 8 °C; when Δt1 = 0, it means that the ambient temperature is the same as the temperature of the GIS device, and the influence of the ambient temperature on the temperature of the GIS device can be ignored. That is, at this time, the Lmax of the GIS device is only affected by the historical load. The same is true when Δt2 = 0, and the historical detection data in which the Lmin of the GIS device is only affected by the historical load can be obtained; find F1 = 36 °C, F2 = 22 °C. Taking the above-mentioned historical detection data as an example, calculate St1 = 35 °C - 36 °C = -1 °C, St2 = 24 °C - 22 °C = 2 °C; Please refer to Figure 4As shown, a plane rectangular coordinate system is established with Δt1 as the horizontal axis and St1 as the vertical axis, named as the maximum temperature influence diagram, and St1 is entered into the maximum temperature influence diagram according to the corresponding Δt1. A plane rectangular coordinate system is established with Δt2 as the horizontal axis and St2 as the vertical axis, named as the minimum temperature influence diagram, and St2 is entered into the minimum temperature influence diagram according to the corresponding Δt2; Perform polynomial regression analysis on the maximum temperature influence diagram and the minimum temperature influence diagram, and name the regression functions as the maximum temperature influence function and the minimum temperature influence function, respectively. This completes the analysis of the restored reference data set. For all historical monitoring data, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively, solve the maximum temperature influence value and the minimum temperature influence value, subtract Lmax from the maximum temperature influence value to obtain the restored value of Lmax, marked as Umax, subtract Lmin from the minimum temperature influence value to obtain the restored value of Lmin, marked as Umin; In practical applications, since the analysis and application processes of the maximum temperature influence diagram and the minimum temperature influence diagram are exactly the same, this embodiment only takes the maximum temperature influence diagram as an example for specific explanation; the maximum temperature influence diagram is constructed as shown in FIG. Figure 4 As shown in the figure, the maximum temperature influence function obtained by polynomial regression analysis is St1=-0.0029×Δt1 2 +0.3142×Δt1-0.404, the maximum temperature influence value can be obtained by solving St1. Taking the historical detection data listed above as an example, Δt1 is -3°C, and the maximum temperature influence value is -1.4°C when substituted into the maximum temperature influence function. The calculation result is retained to one decimal place, and Lmax is subtracted from the maximum temperature influence value to obtain Umax=35°C-(-1.4°C)=36.4°C. The calculation and analysis process of Umin is the same as that of Umax, and no specific description is given in this embodiment. The feature learning unit is used to reanalyze the infrared features of the GIS equipment after restoring the Tmax and Tmin, and then learn the normal infrared features based on the power load and ambient temperature; The feature learning unit is configured with a feature learning strategy, which includes: After restoring Lmax and Lmin in the historical monitoring data, rename the historical monitoring data as restored monitoring data, use Umax in the restored monitoring data as Tmax, and Umin as Tmin, and re-analyze the infrared features of the image to be analyzed; The heat source sets of all restored monitoring data are merged to obtain the characteristic distribution set; In practical applications, the feature analysis module illustrates the extraction process of infrared features. Before extracting the infrared features, it is necessary to first restore Lmax and Lmin. After restoring Lmax and Lmin, the temperatures corresponding to the remaining pixel points can also be calculated and restored to the temperatures unaffected by the ambient temperature through the temperature calculation unit in the feature analysis module, and then the infrared features of the image to be analyzed are extracted; the restored monitoring data also stores the image to be analyzed at the corresponding moment, which is used to extract historical infrared features. The heat source sets of all the restored monitoring data are merged to obtain a feature distribution set. Due to the large amount of data, it is not convenient to show it specifically in this embodiment; Please refer to Figure 5 As shown, a plane rectangular coordinate system is established with the power load as the X-axis and the feature slope as the Y-axis, named the feature distribution diagram. The feature slopes of the restored monitoring data are entered into the feature distribution diagram according to the power load; Perform a linear regression analysis on the feature distribution diagram, name the regression function as the feature distribution function, and at the same time obtain the maximum value of the residuals of the feature distribution function, marked as the fluctuation value; Each monitoring area corresponds to a fluctuation value, a feature distribution function, and a feature distribution set; In practical applications, after extracting the feature slopes of the restored detection data, a feature distribution diagram is constructed as Figure 5 shown. Through polynomial regression analysis, the feature distribution function is obtained as Y = 0.00003×X - 0.0003, and at the same time the fluctuation value is obtained as 0.00013. The fluctuation values, feature distribution functions, and feature distribution sets of each detection area are independent.

[0022] The fault judgment module is used to judge whether there is a thermal fault in the GIS device by combining the power load, ambient temperature, and infrared features; The fault judgment module is configured with a fault judgment strategy, and the fault judgment strategy includes: Real-time monitor the power load and ambient temperature of the substation, and mark them as the real-time load and real-time temperature respectively; For any monitoring area, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively to solve the maximum temperature influence value and the minimum temperature influence value. Subtract the maximum temperature influence value from Lmax to obtain the restored value of Lmax, marked as Vmax. Subtract the minimum temperature influence value from Lmin to obtain the restored value of Lmin, marked as Vmin; Analyze the infrared features of the monitoring area through Vmax and Vmin, including the heat source set and the feature slope, named the real-time set and the real-time slope respectively; In practical applications, take Figure 2Taking the image to be analyzed in the corresponding monitoring area as an example, Hmax and Hmin are 252 and 30 respectively, Tmax and Tmin are 46.2 °C and 22 °C respectively. At the same time, the real-time temperature is monitored to be 26 °C, and the real-time load is 32 MW. It is calculated that Δt1 and Δt2 are -20.2 °C and 4 °C respectively. Substituting Δt1 into the maximum temperature influence function, the maximum temperature influence value is solved to be -7.9 °C. Substituting Δt2 into the minimum temperature influence function, the minimum temperature influence value is solved to be 1 °C. After restoration, Vmax and Vmin are 54.1 °C and 21 °C respectively. Then, the temperature calculation unit calculates Vmax and Vmin, so as to calculate the temperature of each pixel point. Then, the infrared features of the image to be analyzed are extracted to obtain the real-time set and the real-time slope. Among them, the data volume of the real-time set is too large and it is not convenient to display specifically in this embodiment. The real-time slope is 0.00064; Mark the real-time set and the feature distribution set as D1 and D2 respectively. If D1 ∈ D2, output the heat source normal signal, otherwise output the heat source abnormal signal; Substitute the real-time load into the feature distribution function, name the solved value as the ideal feature, increase the ideal feature by the fluctuation value to obtain the maximum feature, and decrease the ideal feature by the fluctuation value to obtain the minimum feature; Judge whether the real-time slope is between the maximum feature and the minimum feature. If so, output the feature normal signal, otherwise output the feature abnormal signal; If the heat source abnormal signal or the feature abnormal signal is output, mark that there is a fault in the corresponding monitoring area; In practical applications, through analysis, it is obtained that D1 ∈ D2, and the heat source normal signal is output. Substitute the real-time load of 32 MW into the feature distribution function, and the solved ideal feature is 0.00066. Then calculate it with the fluctuation value to obtain the maximum feature of 0.00079 and the minimum feature of 0.00053. By looking up, it is found that the ideal feature is between the maximum feature and the minimum feature, and the feature normal signal is output. Since both the heat source normal signal and the feature normal signal are output at the same time, it means Figure 2 There is no fault in the monitoring area shown.

[0023] Embodiment 2. Please refer to Figure 6 As shown, the present application provides a method for detecting the fault location image of a substation GIS device, including the following steps: Step S1, obtain the infrared thermal map of the GIS device of the substation; Step S1 includes the following sub-steps: Step S101, perform real-time monitoring on the GIS device through an infrared thermal imager; Step S102, place the infrared thermal imager in four directions: directly in front of, directly behind, directly to the left of, and directly to the right of the GIS device; Step S103, name the image captured by the infrared thermal imager as the infrared thermal map; Step S2, partition the components of the GIS device in the infrared thermal map to obtain the image to be analyzed; Step S2 includes the following sub-steps: Step S201, the GIS device consists of components with the number of the first component quantity; Step S202, manually select and partition the components in the infrared thermal map by bounding boxes. Each partitioned area is an image to be analyzed, and each image to be analyzed corresponds to a monitoring area; Step S203, manual partitioning only needs to partition the infrared thermal maps in four directions for the first time. When partitioning later, directly obtain the monitoring area corresponding to the first partitioning; Step S3, extract the infrared features of the GIS device based on the image to be analyzed; Step S3 includes the following sub-steps: Step S301, for any image to be analyzed, convert the image to be analyzed into a grayscale image and name it the grayscale image to be analyzed; Step S302, mark the pixel point at the nth row and mth column in the grayscale image to be analyzed as P(n, m), where both n and m are positive integers and (n, m) is the serial number of P; Step S303, obtain the corresponding grayscale value at P(n, m), represented by the symbol H(n, m), find the maximum and minimum values in H(n, m), and mark them as Hmax and Hmin respectively. Obtain the maximum and minimum values of the temperature in the image to be analyzed through the infrared thermal imager, and mark them as Tmax and Tmin respectively; Step S304, calculate the temperature of each P(n, m) through the formula where T(n, m) is the temperature of P(n, m); Step S305, analyze the infrared features of the image to be analyzed based on T(n, m); Step S305 includes the following sub-steps: Step S3051, find the P(n, m) where H(n, m)=Hmax, mark it as the heat source point, and count all the heat source points to obtain the heat source set; Step S3052, sort and number T(n, m) in ascending order, represented by the symbol G i where i is a positive integer and i is the serial number of G; Step S3053, establish a plane rectangular coordinate system with i as the X-axis and G i as the Y-axis, name it the infrared feature map, and input G i into the infrared feature map according to i; Step S3054: Perform a linear regression analysis on the infrared feature map to obtain the slope of the linear regression function, named the feature slope. The heat source set and the feature slope are the infrared features. Step S4: Obtain the power load and ambient temperature in the normal historical monitoring data, restore the temperature of the GIS device in the infrared thermal map through the ambient temperature, and then learn the normal infrared features based on the power load and ambient temperature. Step S4 includes the following sub-steps: Step S401: Perform a restoration analysis on Tmax and Tmin based on the normal historical monitoring data. Step S401 includes the following sub-steps: Step S4011: Analyze any monitoring area to obtain the normal historical monitoring data. The historical monitoring data records the power load, ambient temperature, Tmax, and Tmin at any moment, which are respectively marked as historical load, historical temperature, Lmax, and Lmin. The historical monitoring data of each monitoring area is independent of each other. Step S4012: Find the historical load with the most repeated historical loads, and integrate the corresponding historical monitoring data into a restoration reference data group. Step S4013: Analyze the historical monitoring data in the restoration reference data group, number the historical temperature, and represent it by the symbol R j where j is a positive integer and j is the serial number of R. Mark the corresponding Lmax and Lmin of R j as EA j and EB j respectively. Step S4014: Calculate R j -EA j and R j -EB j , obtain the maximum ambient temperature difference and the minimum of each historical monitoring data, which are respectively marked as Δt1 and Δt2. Find the average value of Lmax when Δt1 is zero, and at the same time find the average value of Lmin when Δt2 is zero, which are respectively marked as F1 and F2. Calculate EA j -F1 to obtain the maximum temperature rise affected by each historical monitoring data, marked as St1. Calculate EA j -F2 to obtain the minimum temperature rise affected by each historical monitoring data, marked as St2. Step S4015: Establish a plane rectangular coordinate system with Δt1 as the horizontal axis and St1 as the vertical axis, named the maximum temperature influence map, and input St1 into the maximum temperature influence map according to the corresponding Δt1. Establish a plane rectangular coordinate system with Δt2 as the horizontal axis and St2 as the vertical axis, named the minimum temperature influence map, and input St2 into the minimum temperature influence map according to the corresponding Δt2. Step S4016, performing polynomial regression analysis on the maximum temperature influence diagram and the minimum temperature influence diagram, and naming the regression functions as the maximum temperature influence function and the minimum temperature influence function, respectively, and thus the analysis of the restored reference data set is completed; Step S4017, for all historical monitoring data, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively, solve the maximum temperature influence value and the minimum temperature influence value, subtract Lmax from the maximum temperature influence value to obtain the restored value of Lmax, marked as Umax, subtract Lmin from the minimum temperature influence value to obtain the restored value of Lmin, marked as Umin; Step S402, after restoring and analyzing Tmax and Tmin, reanalyze the infrared characteristics of the GIS equipment, and then learn the normal infrared characteristics based on the power load and ambient temperature; Step S402 includes the following sub-steps: Step S4021, after restoring Lmax and Lmin in the historical monitoring data, rename the historical monitoring data as restored monitoring data, use Umax in the restored monitoring data as Tmax, and Umin as Tmin, and re-analyze the infrared features of the image to be analyzed; Step S4022, merging all heat source sets of restored monitoring data to obtain a feature distribution set; Step S4023, establish a plane rectangular coordinate system with power load as the X-axis and characteristic slope as the Y-axis, named as characteristic distribution graph, and enter the characteristic slope of the restored monitoring data into the characteristic distribution graph according to the power load; Step S4024, performing linear regression analysis on the characteristic distribution graph, naming the regression function as the characteristic distribution function, and obtaining the maximum value of the residual of the characteristic distribution function, marking it as the fluctuation value; Step S4025, each monitoring area corresponds to a fluctuation value, a characteristic distribution function and a characteristic distribution set; Step S5, judging whether the GIS equipment has a thermal fault by combining the power load, ambient temperature and infrared characteristics; Step S5 includes the following sub-steps: Step S501, real-time monitoring of the power load and ambient temperature of the substation, marked as real-time load and real-time temperature respectively; Step S502: for any monitoring area, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively, solve the maximum temperature influence value and the minimum temperature influence value, subtract Lmax from the maximum temperature influence value to obtain the restored value of Lmax, marked as Vmax, and subtract Lmin from the minimum temperature influence value to obtain the restored value of Lmin, marked as Vmin; Step S503: Analyze the infrared features of the monitoring area through Vmax and Vmin, including the heat source set and the feature slope, which are respectively named the real-time set and the real-time slope. Step S504: Mark the real-time set and the feature distribution set as D1 and D2 respectively. If D1 ∈ D2, output a normal heat source signal; otherwise, output an abnormal heat source signal. Step S505: Substitute the real-time load into the feature distribution function, name the obtained value the ideal feature, increase the ideal feature by the fluctuation value to get the maximum feature, and decrease the ideal feature by the fluctuation value to get the minimum feature. Step S506: Determine whether the real-time slope is between the maximum feature and the minimum feature. If so, output a normal feature signal; otherwise, output an abnormal feature signal. Step S507: If an abnormal heat source signal or an abnormal feature signal is output, mark that there is a fault in the corresponding monitoring area.

[0024] Embodiment 3: The present application provides an electronic device, which may include: a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, the steps in a method for detecting the fault location of a substation GIS device are run to achieve the following functions: obtaining an infrared thermal image of the GIS device of the substation; partitioning the components of the GIS device in the infrared thermal image to obtain an image to be analyzed; extracting the infrared features of the GIS device based on the image to be analyzed; restoring the temperature of the GIS device in the infrared thermal image through the ambient temperature, and then learning the normal infrared features based on the power load and the ambient temperature; combining the power load, the ambient temperature, and the infrared features to determine whether there is a thermal fault in the GIS device.

[0025] In addition, when the logical instructions in the above-mentioned memory are implemented in the form of software function units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. And the aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0026] Embodiment 4. The present application also provides a computer-readable storage medium. The present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method for detecting images of faults in substation GIS equipment are run to achieve the following functions: obtaining an infrared thermal map of the GIS equipment in the substation; partitioning the components of the GIS equipment in the infrared thermal map to obtain an image to be analyzed; extracting the infrared features of the GIS equipment based on the image to be analyzed; restoring the temperature of the GIS equipment in the infrared thermal map through the ambient temperature, and then learning the normal infrared features based on the power load and the ambient temperature; combining the power load, the ambient temperature and the infrared features to determine whether there is a thermal fault in the GIS equipment.

[0027] Through the description of the above embodiments, the embodiments of the present invention can be provided as a method, a system or a computer program product. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0028] In the embodiments provided by the present application, it should be understood that the disclosed system or method can be implemented in other ways. The above-described embodiments are merely illustrative. For example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation. For another example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some communication interfaces. The indirect coupling or communication connection of systems, modules and units can be in an electrical, mechanical or other form.

[0029] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for fault location image detection of substation GIS equipment, characterized in that, It includes the following steps: Obtain the infrared thermal image of the GIS equipment in the substation; Partition the components of the GIS equipment in the infrared thermal image to obtain the image to be analyzed; Extract the infrared features of the GIS equipment based on the image to be analyzed; Obtain the power load and ambient temperature in the normal historical monitoring data, restore the temperature of the GIS equipment in the infrared thermal image through the ambient temperature, and then learn the normal infrared features based on the power load and ambient temperature; Combine the power load, ambient temperature and infrared features to determine whether there is a thermal fault in the GIS equipment.

2. The image detection method for fault location of a substation GIS device according to claim 1, wherein, Obtaining the infrared thermal image of the GIS equipment in the substation includes the following sub-steps: Perform real-time monitoring of the GIS equipment through an infrared thermal imager; The infrared thermal imager is placed in four directions: directly in front of, directly behind, directly to the left, and directly to the right of the GIS equipment; Name the image captured by the infrared thermal imager as the infrared thermal image.

3. The image detection method for fault location of a substation GIS device according to claim 2, wherein, Partitioning the components of the GIS equipment in the infrared thermal image to obtain the image to be analyzed includes the following sub-steps: The GIS equipment is composed of a first number of components; Manually select and partition the components in the infrared thermal image by bounding boxes. Each partitioned area is an image to be analyzed, and each image to be analyzed corresponds to a monitoring area; For manual partitioning, only the infrared thermal images in the four directions need to be partitioned for the first time. When partitioning subsequently, directly obtain the monitoring area corresponding to the first partitioning.

4. A method for detecting images of substation GIS equipment fault location according to claim 3, characterized in that, Extracting the infrared features of the GIS equipment based on the image to be analyzed includes the following sub-steps: For any image to be analyzed, convert the image to be analyzed into a grayscale image and name it the grayscale image to be analyzed; Mark the pixel point at the nth row and mth column in the grayscale image to be analyzed as P(n,m), where both n and m are positive integers and (n,m) is the serial number of P; Obtain the corresponding grayscale value at P(n,m), denoted by the symbol H(n,m), find the maximum and minimum values in H(n,m), and mark them as Hmax and Hmin respectively. Obtain the maximum and minimum values of the temperature in the image to be analyzed through the infrared thermal imager, and mark them as Tmax and Tmin respectively; Calculate the temperature of each P(n,m) through the formula where T(n,m) is the temperature of P(n,m); Analyze the infrared features of the image to be analyzed based on T(n,m).

5. A method for detecting an image of a substation GIS equipment fault location according to claim 4, characterized in that, Analyzing the infrared features of the image to be analyzed based on T(n,m) includes the following sub-steps: Find the P(n,m) where H(n,m)=Hmax and mark it as the heat source point. Count all the heat source points to obtain the heat source set; Sort and number T(n,m) in ascending order, denoted by the symbol G i where i is a positive integer and i is the serial number of G; Taking i as the X-axis and G i as the Y-axis to establish a plane rectangular coordinate system, named the infrared feature map, and input G i into the infrared feature map according to i; Perform linear regression analysis on the infrared feature map to obtain the slope of the linear regression function, named the feature slope. The heat source set and the feature slope are the infrared features.

6. A method for image detection of substation GIS equipment fault location according to claim 5, characterized in that, Obtaining the power load and ambient temperature in the normal historical monitoring data, restoring the temperature of the GIS equipment in the infrared thermal image through the ambient temperature, and then learning the normal infrared features based on the power load and ambient temperature includes the following sub-steps: Perform restoration analysis on Tmax and Tmin based on the normal historical monitoring data; After performing restoration analysis on Tmax and Tmin, re-analyze the infrared features of the GIS equipment, and then learn the normal infrared features based on the power load and ambient temperature.

7. A method for detecting faults in substation GIS equipment by image, according to claim 6, characterized in that The restoration analysis of Tmax and Tmin based on normal historical monitoring data includes the following sub-steps: Analyze any monitoring area to obtain normal historical monitoring data, which records the power load, ambient temperature, Tmax and Tmin at any time, marked as historical load, historical temperature, Lmax and Lmin respectively. The historical monitoring data of each monitoring area are independent of each other; Find the historical load with the most repeated times, and integrate the corresponding historical monitoring data into a restoration reference data group; Analyze the historical monitoring data in the reduction reference data group, number the historical temperatures, and represent them by the symbol R j where j is a positive integer and j is the serial number of R. Mark the Lmax and Lmin corresponding to R j as EA j and EB j respectively; Calculate R j -EA j and R j -EB j to obtain the maximum and minimum ambient temperature differences for each historical monitoring data, which are marked as Δt1 and Δt2 respectively. Find the average value of Lmax when Δt1 is zero, and at the same time find the average value of Lmin when Δt2 is zero, which are marked as F1 and F2 respectively. Calculate EA j -F1 to obtain the maximum temperature rise impact for each historical monitoring data, which is marked as St1. Calculate EA j -F2 to obtain the minimum temperature rise impact for each historical monitoring data, which is marked as St2; A plane rectangular coordinate system is established with Δt1 as the horizontal axis and St1 as the vertical axis, named as the maximum temperature influence diagram, and St1 is recorded in the maximum temperature influence diagram according to the corresponding Δt1; a plane rectangular coordinate system is established with Δt2 as the horizontal axis and St2 as the vertical axis, named as the minimum temperature influence diagram, and St2 is recorded in the minimum temperature influence diagram according to the corresponding Δt2; Perform polynomial regression analysis on the maximum temperature influence diagram and the minimum temperature influence diagram, and name the regression functions as the maximum temperature influence function and the minimum temperature influence function, respectively. This completes the analysis of the restored reference data set. For all historical monitoring data, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively, solve the maximum temperature influence value and the minimum temperature influence value, subtract Lmax from the maximum temperature influence value to obtain the restored value of Lmax, marked as Umax, subtract Lmin from the minimum temperature influence value to obtain the restored value of Lmin, marked as Umin.

8. A method for image detection of substation GIS equipment fault location according to claim 7, characterized in that, After restoring and analyzing Tmax and Tmin, reanalyzing the infrared characteristics of the GIS equipment, and then learning the normal infrared characteristics based on the power load and ambient temperature includes the following sub-steps: After restoring Lmax and Lmin in the historical monitoring data, rename the historical monitoring data as restored monitoring data, use Umax in the restored monitoring data as Tmax, and Umin as Tmin, and re-analyze the infrared features of the image to be analyzed; The heat source sets of all restored monitoring data are merged to obtain the characteristic distribution set; A plane rectangular coordinate system is established with power load as the X-axis and characteristic slope as the Y-axis, named as characteristic distribution diagram, and the characteristic slope of the restored monitoring data is entered into the characteristic distribution diagram according to power load; Perform linear regression analysis on the characteristic distribution graph, name the regression function as the characteristic distribution function, and obtain the maximum value of the residual of the characteristic distribution function, which is marked as the fluctuation value; Each monitoring area corresponds to a fluctuation value, a characteristic distribution function and a characteristic distribution set.

9. A method for detecting an image of a substation GIS equipment fault location according to claim 8, characterized in that, Combining power load, ambient temperature and infrared characteristics, judging whether there is a thermal fault in GIS equipment includes the following sub-steps: Real-time monitoring of the power load and ambient temperature of the substation, marked as real-time load and real-time temperature respectively; For any monitoring area, calculate Δt1 and Δt2 and substitute them into the maximum temperature influence function and the minimum temperature influence function respectively to solve the maximum temperature influence value and the minimum temperature influence value. Subtract the maximum temperature influence value from Lmax to obtain the restored value of Lmax, denoted as Vmax. Subtract the minimum temperature influence value from Lmin to obtain the restored value of Lmin, denoted as Vmin; Analyze the infrared characteristics of the monitoring area through Vmax and Vmin, including the heat source set and the characteristic slope, and name them the real-time set and the real-time slope respectively; Mark the real-time set and the characteristic distribution set as D1 and D2 respectively. If D1 ∈ D2, output the heat source normal signal; otherwise, output the heat source abnormal signal; Substitute the real-time load into the characteristic distribution function, name the obtained value the ideal characteristic, increase the fluctuation value of the ideal characteristic to obtain the maximum characteristic, and decrease the fluctuation value of the ideal characteristic to obtain the minimum characteristic; Judge whether the real-time slope is between the maximum characteristic and the minimum characteristic. If so, output the characteristic normal signal; otherwise, output the characteristic abnormal signal; If the heat source abnormal signal or the characteristic abnormal signal is output, mark that there is a fault in the corresponding monitoring area.

10. A substation GIS equipment fault location image detection system for implementing a substation GIS equipment fault location image detection method according to any one of claims 1-9, characterized in that, It includes an infrared monitoring module, an image partitioning module, a feature analysis module, a feature learning module, and a fault judgment module; the infrared monitoring module, the image partitioning module, the feature analysis module, and the fault judgment module are respectively connected to the feature learning module for data; The infrared monitoring module is used to obtain the infrared thermal image of the GIS equipment in the substation; The image partitioning module is used to partition the components of the GIS equipment in the infrared thermal image to obtain the image to be analyzed; The feature analysis module is used to extract the infrared features of the GIS equipment based on the image to be analyzed; The feature learning module is used to obtain the power load and environmental temperature in the normal historical monitoring data, restore the temperature of the GIS equipment in the infrared thermal image through the environmental temperature, and then learn the normal infrared features based on the power load and environmental temperature; The fault judgment module is used to judge whether there is a thermal fault in the GIS equipment by combining the power load, environmental temperature, and infrared features.

Images