Transformer substation robot inspection system and control method

Abstract

The invention provides a transformer substation robot inspection system and a control method, and relates to the technical field of transformer substation inspection, and the transformer substation robot inspection system comprises a robot body, a control device and a thermal imaging device. The control device is in communication connection with the thermal imaging device and the robot body; the control device is used for acquiring an inspection route and an inspection plan of the transformer substation; the robot body is controlled to inspect the target equipment along the inspection route according to the inspection time, and when the robot body reaches the position of the target equipment, a thermal infrared image of the target equipment is collected through a thermal imaging device; inputting the thermal infrared image into the trained electrical element detection model, and outputting position information corresponding to each electrical element in the thermal infrared image; determining an abnormal electrical element according to the position information corresponding to the electrical element; and generating an abnormal inspection report of the target equipment according to all abnormal electrical elements in the target equipment and the corresponding current temperature data. The inspection efficiency of the transformer substation can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of substation inspection, and in particular to a substation robot inspection system and a control method. Background Art

[0002] In modern society, electricity has become an indispensable energy source for people's production and life. From daily household electricity consumption to large-scale energy demand in industrial production, the stable supply of electricity is the basis for ensuring the normal operation of society. As an important link in realizing voltage conversion, power distribution and control in the power system, the substation is like the "hub" of power transmission. The safe and stable operation of the substation is crucial to ensuring the safety of the power system.

[0003] However, traditional substations mainly rely on manual inspections, and inspectors need to walk to inspect each device in the substation. Therefore, a lot of time and energy are consumed during the inspection process, resulting in low inspection efficiency. Summary of the invention

[0004] The problem solved by the present invention is how to improve the inspection efficiency of transformer substations.

[0005] In order to solve the above problems, the present invention provides a substation robot inspection system and a control method.

[0006] In a first aspect, the present invention provides a substation robot inspection system, comprising a robot body, a control device and a thermal imaging device, wherein the control device and the thermal imaging device are arranged on the robot body, and the control device is respectively connected to the thermal imaging device and the robot body for communication: The control device is used to: Obtaining an inspection route and inspection plan for the substation, wherein the inspection plan includes target equipment to be inspected and a preset inspection time; Controlling the robot body to inspect the target device along the inspection route according to the inspection time, and when reaching the location of the target device, collecting a thermal infrared image of the target device through the thermal imaging device; Input the thermal infrared image into a trained electrical component detection model, and output position information corresponding to each electrical component in the thermal infrared image; Determine the current temperature data of the electrical component according to the position information corresponding to the electrical component; and determine whether the electrical component is an abnormal electrical component according to a comparison result between the acquired historical temperature data of the electrical component and the current temperature data; An abnormal inspection report of the target device is generated according to all abnormal electrical components in the target device and corresponding current temperature data.

[0007] Optionally, inspecting the target device along the inspection route includes: When an obstacle is detected on the inspection route, obtaining boundary position information of the obstacle on the inspection route; generating an obstacle avoidance route according to the boundary position information, wherein a starting point and an end point of the obstacle avoidance route are located on the inspection route; The robot body is controlled to travel from the starting point along the obstacle avoidance route to the end point.

[0008] Optionally, generating an obstacle avoidance route according to the boundary position information includes: Determine the minimum circumscribed rectangle corresponding to the obstacle according to the boundary position information; For each vertex of the minimum circumscribed rectangle, respectively determine the point on the obstacle closest to the vertex, and determine the eccentric distance from the point on the obstacle closest to the vertex to the center point of the minimum circumscribed rectangle; Taking the vertex as the center of the circle and taking the eccentric distance corresponding to the point on the obstacle closest to the vertex as the radius, a vertex circle corresponding to the vertex is generated; Determine a tangent line between two adjacent vertex circles, and determine an intersection point between the tangent line and the corresponding vertex circle, wherein the tangent line does not intersect the minimum circumscribed rectangle; Connecting all the tangent intersection points in sequence to generate a closed route; The starting point and the end point of the obstacle avoidance route are determined according to the intersection of the closed route and the inspection route, and the shortest route from the starting point to the end point in the closed route is used as the obstacle avoidance route.

[0009] Optionally, the training process of the electrical component detection model includes: Acquire historical thermal infrared images of each target device in the substation under different working conditions, and mark the electrical components and corresponding location information in each historical thermal infrared image; Optimizing the grayscale value distribution in the historical thermal infrared image to obtain an optimized historical thermal infrared image; Constructing a training data set based on all the optimized historical thermal infrared images with annotations; The preset initial model is trained according to the training data set to obtain the trained electrical component detection model.

[0010] Optionally, the optimizing according to the gray value distribution in the historical thermal infrared image to obtain the optimized historical thermal infrared image includes: Performing grayscale processing on the historical thermal infrared image to obtain a grayscale value corresponding to each historical pixel point; Determine the number of pixels corresponding to each grayscale value according to the grayscale values ​​of all the historical pixels; According to all the grayscale values ​​and the corresponding number of pixels, determining a replacement grayscale value corresponding to each grayscale value through a grayscale value optimization relationship; The grayscale value of the corresponding historical pixel point is replaced according to the replacement grayscale value to obtain the optimized historical thermal infrared image; The gray value optimization relationship satisfies: ; Among them, I i is the replacement grayscale value corresponding to the i-th grayscale value, H j is the number of pixels corresponding to the jth grayscale value, N is the number of all pixels in the historical thermal infrared image, N min is the number of pixels corresponding to the minimum grayscale value, and i is the number of the first i grayscale values ​​involved in the superposition calculation.

[0011] Optionally, determining the current temperature data of the electrical component according to the position information corresponding to the electrical component includes: Determine the grayscale value corresponding to each pixel of the component in the thermal infrared image according to the position information of the electrical component; Determine the temperature value corresponding to the pixel point of the component according to the gray value corresponding to the pixel point of the component and the preset gray value and temperature matching relationship; Determining the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution of the electrical component according to the temperature values ​​corresponding to all the pixel points of the component; The current temperature data is generated according to the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution.

[0012] Optionally, the current temperature distribution satisfies: Determining whether the pixel point of the component is a high-temperature pixel point according to a comparison result between the temperature value of the pixel point of the component and a preset temperature threshold; Based on a plurality of preset partition center points, determining the partition center line closest to the high-temperature pixel point as the partition center point of the high-temperature pixel point; Generate a corresponding high-temperature partition according to all the high-temperature pixel points corresponding to the same partition center point; The current temperature distribution of the electrical component is generated according to all the high temperature zones.

[0013] Optionally, the current temperature data includes a current average temperature, a current maximum temperature, a current minimum temperature and a current temperature distribution; the historical temperature data includes a historical average temperature, a historical maximum temperature, a historical minimum temperature and a historical temperature distribution; and judging whether the electrical component is an abnormal electrical component according to a comparison result between the current temperature data and the historical temperature data includes: According to the current average temperature, the current maximum temperature, the current minimum temperature, the historical average temperature, the historical maximum temperature and the historical minimum temperature, the temperature deviation of the electrical component is obtained through a preset deviation relationship; When the temperature deviation is greater than a preset deviation threshold, determining that the temperature of the electrical component is abnormal; Determining whether the high temperature area is expanding according to the comparison result of the current temperature distribution and the historical temperature distribution, and if so, determining that the heating area of ​​the electrical component is abnormal; When the temperature of the electrical component is abnormal and / or the heating area is abnormal, determining that the electrical component is the abnormal electrical component; The deviation relationship satisfies: ; Wherein, P is the temperature deviation, is the current average temperature, is the historical average temperature, T max is the current maximum temperature, L max is the historical maximum temperature, T min The current minimum temperature, L min The historical minimum temperature.

[0014] Optionally, judging whether the high temperature area is expanding according to the comparison result between the current temperature distribution and the historical temperature distribution includes: Determine the current high temperature partition area according to all current high temperature partitions in the current temperature distribution, and determine the historical high temperature partition area according to all historical high temperature partitions in the historical temperature distribution; When the current high temperature zone area is larger than the historical high temperature zone area, it is determined that the high temperature region of the electrical component is expanding.

[0015] In a second aspect, the present invention provides a control method for a substation robot inspection system, comprising: Obtaining an inspection route and inspection plan for the substation, wherein the inspection plan includes target equipment to be inspected and a preset inspection time; Controlling the robot body to inspect the target device along the inspection route according to the inspection time, and when reaching the location of the target device, collecting a thermal infrared image of the target device through a thermal imaging device; Input the thermal infrared image into a trained electrical component detection model, and output position information corresponding to each electrical component in the thermal infrared image; Determine the current temperature data of the electrical component according to the position information corresponding to the electrical component; and determine whether the electrical component is an abnormal electrical component according to a comparison result between the acquired historical temperature data of the electrical component and the current temperature data; An abnormal inspection report of the target device is generated according to all abnormal electrical components in the target device and corresponding current temperature data.

[0016] The beneficial effects of the substation robot inspection system and control method of the present invention are as follows: the robot body automatically inspects the target equipment according to the preset inspection route and time, which can avoid the consumption of manpower and time generated by manual inspection, and effectively improve the speed and efficiency of the inspection. The thermal infrared image of the target equipment is collected by the thermal imaging device on the robot body, and the current temperature distribution of the target equipment can be obtained through the thermal infrared image, so as to judge the heating state of the electrical component by the temperature distribution. The thermal infrared image collected by the thermal imaging device is input into the trained electrical component detection model to obtain the corresponding position information of each electrical component in the target equipment in the thermal infrared image, and the accurate temperature data of the electrical component can be quickly obtained by the corresponding position information of the electrical component, which effectively improves the efficiency of collecting the temperature data of the electrical components in the target equipment, and avoids the omissions and errors that may be caused by manual detection. By comparing the current temperature data with the historical temperature data, it is possible to accurately judge whether the heating of the electrical component is abnormal, determine the electrical component with abnormal heating as an abnormal electrical component, and screen the electrical components that may have potential failures by temperature changes, so as to reduce the loss and risk caused by the failure of the target equipment by timely processing the abnormal equipment. Finally, an abnormal inspection report of the target device is generated based on all abnormal electrical components in the target device and the corresponding current temperature data, so that the data information of all abnormally heated electrical components in the target device can be obtained through the abnormal inspection report. The substation robot inspection system can inspect the target equipment according to the set inspection route and inspection time, avoiding the time waste that may be caused by manual inspection. At the same time, it is not limited by working hours and can be inspected according to the inspection plan around the clock, greatly improving the efficiency of substation inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of the structure of a substation robot inspection system according to an embodiment of the present invention; Figure 2 A schematic diagram of an obstacle avoidance route according to an embodiment of the present invention; Figure 3The present invention is a flowchart of another substation robot inspection system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be interpreted as being limited to the embodiments described herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not intended to limit the scope of protection of the present invention.

[0019] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0020] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to"; the term "based on" means "based at least in part on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optionally" means "optional embodiments". The relevant definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc. mentioned in the present invention are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0021] It should be noted that the modifications of "one" and "plurality" mentioned in the present invention are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0022] The names of the messages or information exchanged between multiple devices in the embodiments of the present invention are only used for illustrative purposes, and are not used to limit the scope of these messages or information.

[0023] like Figure 1 As shown, an embodiment of the present invention provides a substation robot inspection system, including a robot body, a control device and a thermal imaging device, wherein the control device and the thermal imaging device are arranged on the robot body, and the control device is respectively communicated with the thermal imaging device and the robot body.

[0024] Specifically, in the substation robot inspection system, the robot body, as the mobile carrier of the inspection system, can move flexibly in the substation according to the preset inspection route, and can move by wheels, tracks, etc., to adapt to different site conditions, such as flat roads, steps and slopes. The thermal imaging device is set at a suitable position on the robot body, so that the thermal imaging device can accurately collect thermal infrared images of the target equipment. The thermal imaging device is mainly used to detect the temperature distribution of the equipment in the substation. Without direct contact with the equipment, the temperature of the equipment surface can be quickly and accurately detected, which can avoid damage to the equipment and contact risks, and improve the safety of the inspection. At the same time, a communication device is set on the robot body. The communication device is used for communication between the inspection system and the remote monitoring center. The communication device receives instructions sent by the monitoring center and obtains data information, such as starting inspection, stopping inspection and adjusting parameters. The communication device transmits the instructions and data information of the remote monitoring center to the control device, which performs corresponding operations and data processing, and sends the inspection results to the remote monitoring center through the communication device. It can also enable the monitoring center to remotely monitor the operating status of the inspection system, including the position, power and equipment status of the robot body, so that system failures or abnormal conditions can be discovered in time and corresponding measures can be taken to deal with them. The control device, thermal imaging device and communication device are arranged on the robot body, and the control device is respectively connected to the thermal imaging device, the communication device and the robot body, so that the thermal imaging device, the communication device and the robot body can be accurately controlled by signals to execute corresponding commands.

[0025] The control device is used to: Obtain an inspection route and inspection plan for the substation, wherein the inspection plan includes target equipment to be inspected and a preset inspection time.

[0026] Specifically, the control device receives the inspection instruction information issued by the remote monitoring center through the communication device, that is, the pre-set inspection route and inspection plan, wherein the inspection route of the robot is determined according to the layout of the substation and the location of the target equipment to be inspected, so that the robot can smoothly reach the location of the target equipment along the inspection route. The inspection target equipment in the inspection plan will cover various key equipment that need to be inspected in the substation, such as transformers, switch cabinets, transformers and capacitor banks. The preset inspection time can set a fixed inspection time, such as inspecting the target equipment once every 3 hours along the inspection route. It can also be determined according to the importance of the equipment, the operating status and relevant regulations. The inspection frequency can be increased for important or fault-prone equipment, such as the main transformer, which may need to be inspected once every hour; while for some relatively stable equipment, such as some auxiliary equipment, the inspection interval may be slightly longer, such as once a day or once a week. The substation robot inspection system will not be affected by human factors such as fatigue and inattention, and can always maintain stable detection accuracy, providing more accurate and reliable data. In addition, there are certain risks in manual inspections of some dangerous areas or high-voltage equipment, and the substation robot inspection system can prevent personnel from being directly exposed to dangerous environments, reducing the occurrence of safety accidents.

[0027] The robot body is controlled to inspect the target device along the inspection route according to the inspection time, and when the robot body reaches the position of the target device, a thermal infrared image of the target device is collected by the thermal imaging device.

[0028] Specifically, the control device controls the robot body to perform inspections along a set inspection route according to a preset inspection time, that is, to reach the location of the target device along the set inspection route. When arriving at the location of the target device, the control device controls the thermal imaging device located at the front end of the robot body to capture a thermal infrared image of the target device. The thermal imaging device can also be set on the mechanical arm of the robot body, so that the thermal imaging device can capture a complete thermal infrared image of the target device by controlling the mechanical arm.

[0029] The thermal infrared image is input into a trained electrical component detection model, and the position information corresponding to each electrical component in the thermal infrared image is output.

[0030] Specifically, the temperature distribution of electrical components in the target device can be reflected through the collected thermal infrared image, so that the working status and potential faults of the electrical components can be judged. The collected thermal infrared image is input into the trained electrical component detection model, and each electrical component identified from the thermal infrared image and the specific location information of the electrical component in the thermal infrared image are output. The trained electrical component detection model is obtained by training the initial model based on a large amount of thermal infrared image data and related annotation information. The initial model can be constructed by convolutional neural network, recurrent neural network or adversarial neural network.

[0031] The current temperature data of the electrical component is determined according to the position information corresponding to the electrical component; and whether the electrical component is an abnormal electrical component is determined according to a comparison result between the acquired historical temperature data of the electrical component and the current temperature data.

[0032] Specifically, based on the position information of each electrical component in the thermal infrared image, the pixel point information of the corresponding area of ​​the electrical component in the thermal infrared image is extracted, and for each pixel in the extracted pixel point information, the current temperature data of the corresponding area of ​​the electrical component is calculated through the correspondence between the grayscale value and the temperature, wherein the current temperature data may include data information such as the average value, maximum value, minimum value and temperature distribution, so that the temperature condition of the electrical component at the current moment can be fully understood.

[0033] It should be noted that when determining the correspondence between grayscale values ​​and temperature, the thermal infrared imaging device needs to be calibrated. During the calibration process, thermal infrared images will be acquired under different known temperature conditions and the corresponding grayscale values ​​will be recorded. Through multiple corresponding points of temperature and grayscale values, a mapping relationship or calibration curve between grayscale value and temperature can be established, that is, the correspondence between grayscale value and temperature.

[0034] Furthermore, when it is necessary to compare with the temperature data at historical moments, the control device obtains the historical temperature data corresponding to the electrical component from the remote control center through the communication device, wherein the historical temperature data is the historical temperature data of the electrical component at the same historical time node as the current electrical component operation time and operation state. The temperature distribution state of the electrical component under the historical working state matching the current moment can be obtained through the historical temperature data. Since the historical temperature data includes the temperature fluctuations under different time periods and different load conditions, the historical temperature data obtained at the current moment is the historical temperature data matching the electrical component with the current moment and load state. For example, the current moment is 8 pm, and the load rate of the transformer is 80%. When obtaining the historical temperature data of the transformer, it is necessary to select the historical temperature data corresponding to the time close and the load rate of 80%, so that the current temperature data obtained at the current moment can be more accurately judged for the heating situation through the matched historical temperature data.

[0035] Specifically, the current temperature data is compared with the acquired historical temperature data. The comparison aspects may include the difference between the current average temperature and the historical average temperature value, the difference between the current maximum temperature and the historical maximum temperature, the difference between the current minimum temperature and the historical minimum temperature, and the consistency between the current temperature distribution and the historical temperature distribution. When the temperature difference or temperature distribution exceeds the acceptable set range, it means that the electrical component is abnormally heated, that is, it is an abnormal electrical component. When the deviations of various temperature differences and distribution conditions are not large and within a reasonable controllable range, they will not affect the operation of the electrical component, which means that the electrical component is operating normally and there is no abnormal heating. If the current temperature data obviously exceeds the normal range of the historical temperature, for example, the current temperature is too high or too low, it may indicate that the electrical component is abnormal.

[0036] An abnormal inspection report of the target device is generated according to all abnormal electrical components in the target device and corresponding current temperature data.

[0037] Specifically, the target device may include both electrical components with normal heating and electrical components with abnormal heating. When generating an abnormal inspection report for the target device, the names of all abnormal electrical components and the corresponding current temperature data are recorded, that is, the temperature data information of all electrical components with abnormal heating is recorded in the abnormal inspection report. Therefore, the heating conditions of the electrical components with abnormal heating can be accurately understood through the abnormal inspection report of the target device. The technicians can further inspect the abnormal inspection report of the target device to determine the cause of the abnormal heating, and carry out timely processing and maintenance, thereby ensuring the safe and stable operation of the electrical equipment.

[0038] In this embodiment, the robot body automatically inspects the target device according to the preset inspection route and time, which can avoid the consumption of manpower and time generated by manual inspection, and effectively improve the speed and efficiency of the inspection. The thermal infrared image of the target device is collected by the thermal imaging device on the robot body, and the current temperature distribution of the target device can be obtained through the thermal infrared image, so as to judge the heating state of the electrical component by the temperature distribution. The thermal infrared image collected by the thermal imaging device is input into the trained electrical component detection model to obtain the corresponding position information of each electrical component in the target device in the thermal infrared image. The accurate temperature data of the electrical component can be quickly obtained through the corresponding position information of the electrical component, which effectively improves the efficiency of collecting the temperature data of the electrical component in the target device, and avoids the omissions and errors that may be caused by manual inspection. By comparing the current temperature data with the historical temperature data, it is possible to accurately judge whether the heating of the electrical component is abnormal, determine the electrical component with abnormal heating as an abnormal electrical component, and screen the electrical components that may have potential failures through temperature changes, so as to reduce the losses and risks caused by the failure of the target device by timely processing the abnormal equipment. Finally, an abnormal inspection report of the target device is generated based on all abnormal electrical components in the target device and the corresponding current temperature data, so that the data information of all abnormally heated electrical components in the target device can be obtained through the abnormal inspection report. The substation robot inspection system can inspect the target equipment according to the set inspection route and inspection time, avoiding the time waste that may be caused by manual inspection. At the same time, it is not limited by working hours and can be inspected according to the inspection plan around the clock, greatly improving the efficiency of substation inspection.

[0039] Optionally, inspecting the target device along the inspection route includes: When an obstacle is detected on the inspection route, obtaining boundary position information of the obstacle on the inspection route; generating an obstacle avoidance route according to the boundary position information, wherein a starting point and an end point of the obstacle avoidance route are located on the inspection route; The robot body is controlled to travel from the starting point along the obstacle avoidance route to the end point.

[0040] Specifically, the robot body can be equipped with a variety of sensors to detect obstacles on the inspection route. For example, laser radar can be used to construct a three-dimensional map of the surrounding environment by emitting laser beams and measuring the time of reflected light. When the reflected signal is detected to be inconsistent with the expected open space, it may be judged that there is an obstacle. Ultrasonic sensors can also be used to emit ultrasonic waves and receive echoes. According to the echo time and intensity, it can be judged whether there is an object blocking the front. Similarly, cameras can also be used for obstacle detection. Through image recognition algorithms, objects in the image can be identified and judged whether they are obstacles. In order to improve the accuracy and reliability of obstacle detection, data from multiple sensors can be fused. For example, the distance information of laser radar and the image information of the camera can be combined, and the approximate position and distance of the object can be determined by using laser radar, and the type and specific characteristics of the object can be further determined through the image recognition of the camera.

[0041] Furthermore, when an obstacle is detected, it is necessary to determine the boundary position information of the obstacle. For obstacles detected by the laser radar, the boundary of the obstacle can be determined based on the position of the reflection point of the laser beam. For example, if laser beams emitted from multiple directions are reflected back in a certain area, then the boundary of this area can be determined as the boundary of the obstacle, and the coordinate information corresponding to the boundary of the obstacle is determined in the coordinate system where the inspection route is located, that is, the boundary position information. For obstacles detected by the camera, the boundary pixel position of the object in the image can be determined by the image analysis algorithm, and then the boundary in the image can be converted into position information in the coordinate system, that is, the boundary position information, by combining the parameters of the camera and the coordinate system where the inspection route is located. According to the boundary information, a path planning algorithm can be used to generate an obstacle avoidance route, wherein the starting point and end point of the obstacle avoidance route must be located on the inspection route to ensure that the robot can return to the inspection route to continue to perform the task after avoiding the obstacle. The starting point is usually the position of the robot when it detects the obstacle, or the intersection of the obstacle avoidance route and the inspection route close to the current position of the robot, and the end point is the point where the robot returns to the inspection route after bypassing the obstacle along the obstacle avoidance route, that is, another added point between the obstacle avoidance route and the inspection route. The obstacle avoidance route can be determined by algorithms such as greedy algorithms, simulated annealing algorithms, and A* algorithms. The above algorithms are all prior art and will not be repeated here. After the obstacle avoidance route is generated, the robot body is controlled to reach the starting point of the obstacle avoidance route located on the inspection route, and after driving along the obstacle avoidance route to bypass the obstacle, it reaches the end point of the obstacle avoidance route located on the inspection route. After reaching the end point, the robot body continues to inspect according to the preset inspection route.

[0042] In this optional embodiment, when an obstacle is detected on the inspection route, the robot can automatically obtain the obstacle boundary position information and generate an obstacle avoidance route, greatly reducing the inspection interruption time caused by the obstacle. The robot can quickly adjust the route and continue to perform the inspection task, improving the overall inspection efficiency.

[0043] Optionally, generating an obstacle avoidance route according to the boundary position information includes: Determine the minimum circumscribed rectangle corresponding to the obstacle according to the boundary position information; For each vertex of the minimum circumscribed rectangle, respectively determine the point on the obstacle closest to the vertex, and determine the eccentric distance from the point on the obstacle closest to the vertex to the center point of the minimum circumscribed rectangle; Taking the vertex as the center of the circle and taking the eccentric distance corresponding to the point on the obstacle closest to the vertex as the radius, a vertex circle corresponding to the vertex is generated; Determine a tangent line between two adjacent vertex circles, and determine an intersection point between the tangent line and the corresponding vertex circle, wherein the tangent line does not intersect the minimum circumscribed rectangle; Connecting all the tangent intersection points in sequence to generate a closed route; The starting point and the end point of the obstacle avoidance route are determined according to the intersection of the closed route and the inspection route, and the shortest route from the starting point to the end point in the closed route is used as the obstacle avoidance route.

[0044] Specifically, for a given obstacle boundary position information, that is, the position coordinates of the obstacle boundary in the inspection route coordinate system, a minimum rectangle that can completely surround the obstacle area can be found by calculation. This rectangle has the smallest area, while ensuring that all parts of the obstacle are inside or on the boundary of the rectangle, wherein the determination of the minimum circumscribed rectangle is usually based on geometric algorithms, such as the rotating caliper algorithm, the brute force enumeration method, or the principal axis-based method, etc. The above-mentioned geometric algorithms are all prior art and will not be repeated here. Through the positional relationship of each point on the obstacle boundary, the minimum rectangle that can cover all points is found. The minimum circumscribed rectangle can provide a simplified representation of the obstacle shape, which is convenient for subsequent analysis and processing, thereby helping to determine the relative positional relationship between the robot and the obstacle, as well as the reference boundary when planning the obstacle avoidance route.

[0045] Further, for each vertex of the minimum circumscribed rectangle, it is necessary to find the point on the obstacle that is closest to the vertex. By traversing the boundary position information of the obstacle, the distance from the vertex to each point on the obstacle is calculated, and the point with the smallest distance is found as the nearest point corresponding to the vertex. The geometric distance calculation formula, such as Euclidean distance, Manhattan distance and Chebyshev distance, can be used to calculate the distance between the vertex and the boundary point of the obstacle. The above algorithms are all prior art and will not be repeated here. After determining the vertex and the corresponding nearest point, the distance from this nearest point to the center point of the minimum circumscribed rectangle is calculated, that is, the centrifugal distance. The vertex of the minimum circumscribed rectangle is taken as the center of the circle, and the centrifugal distance corresponding to the point closest to the vertex is taken as the radius to generate the vertex circle corresponding to the vertex. The vertex circle generated in this way can tightly surround the obstacle and provide a reference for subsequent path planning. The center point of the minimum circumscribed rectangle can be calculated by the coordinates of the four vertices of the rectangle, for example, taking the average value of the coordinates of the two diagonal vertices of the rectangle. Since the minimum rectangle has four vertices, the vertex circles corresponding to the four vertices can be obtained respectively by the above method. The vertex circle provides a safe buffer for the robot to ensure that it will not get too close to the obstacle during the obstacle avoidance process. At the same time, these circles also provide the basis for generating tangent lines and closed routes.

[0046] Furthermore, for two adjacent vertex circles, the tangent between them is found through geometric calculation. The generation of the tangent must meet the condition of not intersecting with the minimum circumscribed rectangle, so as to ensure that the obstacle avoidance route is outside the obstacle. The analytic geometry method can be used to determine the equation of the tangent according to the equation of the circle and the distance formula between the two points, so as to obtain the intersection points of the tangent and the corresponding vertex circle. These intersection points will become the key nodes for constructing a closed route. Among them, the coordinates of the intersection points can be obtained by solving the equation of the tangent and the equation of the circle simultaneously. Connect all the tangent intersection points in sequence to form a closed route around the obstacle. This closed route is completely outside the obstacle, providing a safe obstacle avoidance path for the robot. When connecting the intersection points, the continuity and integrity of the route must be ensured. The starting point and end point of the obstacle avoidance route are determined according to the intersection point of the closed route and the inspection route. The starting point can be the position of the robot on the inspection route when encountering an obstacle, and the end point can be the position of the robot returning to the inspection route after avoiding the obstacle. When determining the obstacle avoidance route, the shortest route from the starting point along the closed route to the end point is selected as the final obstacle avoidance route.

[0047] For example, Figure 2 R is the minimum circumscribed rectangle calculated based on the obstacle edge position information, where the four vertices are P1, P2, P3 and P4, respectively. The four vertices are used as the center of the circle, and the eccentric distance corresponding to the point with the minimum distance from each vertex is used as the radius. Figure 2For the four vertex circles shown in the figure, two adjacent circles are used as tangents to obtain the tangent points between the tangent and the two adjacent circles. For example, two adjacent circles with P1 and P4 as centers can obtain two tangent points Q1 and Q8 on the premise that the tangent does not pass through the minimum inscribed rectangle R, where the tangent is the line connecting Q1 and Q8. Similarly, the tangent points Q2 and Q3 of the two adjacent circles with P1 and P2 as centers, the tangent points Q4 and Q5 of the two adjacent circles with P2 and P3 as centers, and the tangent points Q6 and Q7 of the two adjacent circles with P3 and P4 as centers can be obtained. Q7, connect the tangent points Q1, Q2, Q3, Q4, Q5, Q6, Q7 and Q8 in sequence to obtain a closed route S, where the intersection points of the inspection route L and the closed route S are the starting point A and the end point B respectively. According to the starting point A, the end point B and the closed route S, two routes to bypass obstacles can be obtained, that is, route 1 is A-Q4-Q3-Q2-Q1-B, and route 2 is A-Q5-Q6-Q7-Q8-B. After calculation, it can be seen that the driving distance of route 2 is less than the driving distance of route 1, so route 2 is used as the final obstacle avoidance route.

[0048] In this optional embodiment, the obstacle avoidance route generated can ensure that the inspection task is carried out uninterruptedly, improve the adaptability to the load environment, avoid collisions and ensure stable driving, effectively improving the safety and efficiency of the inspection.

[0049] Optionally, the training process of the electrical component detection model includes: Acquire historical thermal infrared images of each target device in the substation under different working conditions, and mark the electrical components and corresponding location information in each historical thermal infrared image; Optimizing the grayscale value distribution in the historical thermal infrared image to obtain an optimized historical thermal infrared image; Constructing a training data set based on all the optimized historical thermal infrared images with annotations; The preset initial model is trained according to the training data set to obtain the trained electrical component detection model.

[0050] Specifically, the historical thermal infrared images of each target device in the substation under different working conditions are obtained, such as thermal infrared images collected at different times and under different operating conditions. These images reflect the thermal radiation of each device in the substation under various normal and possible abnormal working conditions. At the same time, the electrical components in each historical thermal infrared image are annotated in detail, including the type (such as transformer, circuit breaker, etc.) and precise location information of the electrical components. For the acquired historical thermal infrared images, they are optimized according to the distribution characteristics of the grayscale values, so as to obtain optimized historical thermal infrared images. The optimization process involves image enhancement, denoising, contrast adjustment and other processing, so that the thermal characteristics of the electrical components can be presented more clearly through optimization. All the optimized thermal infrared images with accurate annotations are used to generate a training data set, which contains rich heating data information of electrical components, thereby providing rich samples for model training. The preset initial model is trained using the training data set, wherein the initial model can be constructed by convolutional neural network, regional convolutional neural network, recurrent neural network, etc. During the training process, the model learns the characteristic information of electrical components of different types and positions under various working conditions. By continuously adjusting the parameters of the model, it can accurately identify and locate electrical components in newly input thermal infrared images. Therefore, the trained electrical component detection model can accurately identify the electrical components and position information in the collected thermal infrared images, providing reliable data support for actual inspection work.

[0051] In this optional embodiment, by acquiring a large number of historical thermal infrared images under different working conditions and optimizing them, the model can learn more comprehensive and accurate thermal characteristics of electrical components, thereby improving the accuracy of electrical component detection.

[0052] Optionally, the optimizing according to the gray value distribution in the historical thermal infrared image to obtain the optimized historical thermal infrared image includes: Performing grayscale processing on the historical thermal infrared image to obtain a grayscale value corresponding to each historical pixel point; Determine the number of pixels corresponding to each grayscale value according to the grayscale values ​​of all the historical pixels; According to the number of pixels corresponding to all the gray value pairs, determining a replacement gray value corresponding to each gray value through a gray value optimization relationship; The optimized historical thermal infrared image is obtained by replacing the grayscale value of the historical pixel point corresponding to the replacement grayscale value.

[0053] Optionally, the gray value optimization relationship satisfies: ; Among them, I iis the replacement grayscale value corresponding to the i-th grayscale value, H j is the number of pixels corresponding to the jth grayscale value, N is the number of all pixels in the historical thermal infrared image, N min is the number of pixels corresponding to the minimum grayscale value, and i is the number of the first i grayscale values ​​involved in the superposition calculation.

[0054] Specifically, the historical thermal infrared image is gray-processed to obtain the corresponding gray-scale historical thermal infrared image. In the gray-scale image, each pixel has only one gray-scale value. This processing can simplify the image information and focus more on the intensity distribution of thermal radiation. The gray-scale values ​​of all pixels are counted to determine the number of pixels corresponding to the same gray-scale value, so that the distribution of different gray-scale values ​​in the image can be understood, such as which gray-scale values ​​appear more frequently and which ones appear less frequently. According to all gray-scale values ​​and the corresponding number of pixels, the replacement gray-scale value corresponding to each gray-scale value is calculated through the set gray-scale value optimization relationship. The purpose is to adjust the current gray-scale value to better highlight the key information in the image and enhance the contrast and clarity of the image. The corresponding gray-scale value in the historical thermal infrared image is replaced with the calculated replacement gray-scale value to obtain an optimized historical thermal infrared image. The image retains the original thermal radiation information and can be optimized to more clearly display the thermal characteristics of electrical components, which is conducive to subsequent labeling and model training.

[0055] Optionally, determining the current temperature data of the electrical component according to the position information corresponding to the electrical component includes: Determine the grayscale value corresponding to each pixel of the component in the thermal infrared image according to the position information of the electrical component; Determine the temperature value corresponding to the pixel point of the component according to the gray value corresponding to the pixel point of the component and the preset gray value and temperature matching relationship; Determining the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution of the electrical component according to the temperature values ​​corresponding to all the pixel points of the component; The current temperature data is generated according to the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution.

[0056] Specifically, the position information of the electrical component in the thermal infrared image is used to accurately locate the pixel area corresponding to the electrical component in the thermal infrared image, and the grayscale value corresponding to each pixel point belonging to the electrical component is determined. These grayscale values ​​can reflect the thermal radiation intensity of the area represented by the pixel point. Through the pre-set grayscale and temperature matching relationship, the grayscale value of each component pixel point can be converted into a corresponding temperature value, wherein the grayscale value and temperature matching relationship is usually established during the calibration process of the thermal imaging device, ensuring the accurate conversion from grayscale value to temperature value. The temperature values ​​corresponding to all component pixels are processed to obtain the corresponding current average temperature, which reflects the average thermal state of the electrical component as a whole; the current maximum temperature and the current minimum temperature, these two values ​​can help identify the hottest and coldest areas in the component, which are very important for discovering potential hot spots and abnormal cooling areas; at the same time, the current temperature distribution is analyzed to understand the temperature changes on the entire electrical component. Finally, the current temperature data of the electrical component is generated according to the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution. The current temperature state of the electrical component is fully described through the current temperature data, which provides an important basis for subsequent analysis and judgment.

[0057] Exemplarily, the current average temperature satisfies: ; in, is the current average temperature, M is the number of all component pixels, T k is the temperature value corresponding to the kth component pixel.

[0058] Optionally, the current temperature distribution satisfies: Determining whether the pixel point of the component is a high-temperature pixel point according to a comparison result between the temperature values ​​of all the pixel points of the component and a preset temperature threshold; Based on a plurality of preset partition center points, determining the partition center line closest to the high-temperature pixel point as the partition center point of the high-temperature pixel point; Generate a corresponding high-temperature partition according to all the high-temperature pixel points corresponding to the same partition center point; The current temperature distribution of the electrical component is generated according to all the high temperature zones.

[0059] Specifically, the temperature value corresponding to each component pixel is compared with a preset temperature threshold. If the temperature value of a component pixel is higher than the temperature threshold, the pixel is judged to be a high-temperature pixel, wherein the temperature threshold is usually set according to the normal working temperature range of the electrical component and safety considerations. Exceeding this threshold may mean that there is abnormal heating in the area. Further, multiple partition center points are preset, and the distance from each partition center point is calculated according to the position information of each pixel determined to be a high-temperature pixel, and the partition center point with the closest distance is determined as the partition center point corresponding to the high-temperature pixel point, so as to classify the high-temperature pixels to better understand the distribution of high-temperature areas. According to all high-temperature pixels corresponding to the same partition center point, the corresponding high-temperature partition is generated, that is, the high-temperature pixels belonging to the same partition center point are classified into a high-temperature partition, which represents a specific high-temperature area on the electrical component. By combining all high-temperature partitions, the current temperature distribution of the electrical component can be generated. Through this temperature distribution, it can be clearly shown which areas on the electrical component have higher temperatures, as well as the specific location and range of the high-temperature area, which provides an important basis for further analyzing the working status and potential faults of the electrical component.

[0060] Optionally, the current temperature data includes a current average temperature, a current maximum temperature, a current minimum temperature and a current temperature distribution; the historical temperature data includes a historical average temperature, a historical maximum temperature, a historical minimum temperature and a historical temperature distribution; and judging whether the electrical component is an abnormal electrical component according to a comparison result between the current temperature data and the historical temperature data includes: According to the current average temperature, the current maximum temperature, the current minimum temperature, the historical average temperature, the historical maximum temperature and the historical minimum temperature, the temperature deviation of the electrical component is obtained through a preset deviation relationship; When the temperature deviation is greater than a preset deviation threshold, determining that the temperature of the electrical component is abnormal; Determining whether the high temperature area is expanding according to the comparison result of the current temperature distribution and the historical temperature distribution, and if so, determining that the heating area of ​​the electrical component is abnormal; When the temperature of the electrical component is abnormal and / or the heating area is abnormal, the electrical component is determined to be the abnormal electrical component.

[0061] Optionally, the deviation relationship satisfies: ; Wherein, P is the temperature deviation, is the current average temperature, is the historical average temperature, T max is the current maximum temperature, L maxis the historical maximum temperature, T min The current minimum temperature, L min The historical minimum temperature.

[0062] Specifically, the temperature deviation of the electrical component is calculated based on the current average temperature, the current maximum temperature, the current minimum temperature, the historical average temperature, the historical maximum temperature and the historical minimum temperature through a preset deviation relationship. When the calculated temperature deviation is greater than the preset deviation threshold, the temperature of the electrical component is judged to be abnormal. When the temperature deviation is less than or equal to the deviation threshold, the temperature of the electrical component is judged to be normal. The deviation threshold can be pre-set based on factors such as the normal operating temperature fluctuation range of the electrical component and safety requirements. If the temperature deviation exceeds this range, it means that the current temperature has changed significantly compared with the historical temperature, and the electrical component is in an abnormal state, and there may be a risk of overheating.

[0063] Furthermore, the current temperature distribution is compared with the historical temperature distribution. The current temperature distribution reflects the temperature conditions of each area on the current electrical component, while the historical temperature distribution is the temperature distribution of the time nodes in the past that match the device operation time and operation status. By comparing, the change of the high-temperature area in the electrical component can be judged, that is, whether it has expanded. The current high-temperature area is larger than the historical high-temperature area, indicating that the heating condition of the electrical component has changed, and there may be abnormal heating diffusion caused by local overheating or faults. When the electrical component has abnormal temperature and / or abnormal heating area, the electrical component is judged to be an abnormal electrical component, that is, as long as one of the abnormal conditions or both conditions are met at the same time, the electrical component can be determined to be in an abnormal state and requires further inspection and maintenance.

[0064] In this optional embodiment, by comprehensively considering multiple aspects of current temperature data and historical temperature data, including average temperature, maximum temperature, minimum temperature, and temperature distribution, it is possible to more accurately determine whether an electrical component is an abnormal electrical component, providing an important basis for the inspection and maintenance of the substation.

[0065] Optionally, judging whether the high temperature area is expanding according to the comparison result between the current temperature distribution and the historical temperature distribution includes: Determine the current high temperature partition area according to all current high temperature partitions in the current temperature distribution, and determine the historical high temperature partition area according to all historical high temperature partitions in the historical temperature distribution; When the current high temperature zone area is larger than the historical high temperature zone area, it is determined that the high temperature region of the electrical component is expanding.

[0066] Specifically, all current high-temperature partitions are obtained from the current temperature distribution, which describes the temperature conditions of various regions of the electrical component in the current state, wherein the high-temperature partition refers to the region where the temperature exceeds a certain threshold. For each current high-temperature partition, the corresponding area can be determined, and its area can be approximately represented by calculating the number of pixels contained in the high-temperature partition. All historical high-temperature partitions are found from the historical temperature partitions. The historical temperature partitions reflect the temperature distribution of the electrical component at the time nodes that matched the current electrical component operation state in the past. The area of ​​each historical high-temperature partition is calculated using the same method as that for determining the area of ​​the current high-temperature partition, that is, by the number of pixels or the actual area value after conversion. After the current high-temperature partition area and the historical high-temperature partition area are determined, the two are compared. If the current high-temperature partition area is larger than the historical high-temperature partition area, it means that in the current state, the high-temperature area of ​​the electrical component has expanded compared to the past, which may mean that the heating condition of the electrical component has changed, and there may be potential faults or abnormal conditions.

[0067] In this optional embodiment, by comparing the high-temperature partition areas in the current temperature distribution and the historical temperature distribution, it is possible to effectively determine whether the high-temperature area of ​​the electrical component has expanded, providing an important basis for further analyzing the working status and potential problems of the electrical component.

[0068] like Figure 3 As shown, a control method of a substation robot inspection system provided by an embodiment of the present invention includes: S100, obtaining an inspection route and an inspection plan of the substation, wherein the inspection plan includes target equipment to be inspected and a preset inspection time; S200, controlling the robot body to inspect the target device along the inspection route according to the inspection time, and when reaching the location of the target device, collecting a thermal infrared image of the target device through a thermal imaging device; S300, inputting the thermal infrared image into a trained electrical component detection model, and outputting position information corresponding to each electrical component in the thermal infrared image; S400, determining current temperature data of the electrical component according to the position information corresponding to the electrical component; and judging whether the electrical component is an abnormal electrical component according to a comparison result between the acquired historical temperature data of the electrical component and the current temperature data; S500: Generate an abnormal inspection report for the target device according to all abnormal electrical components in the target device and corresponding current temperature data.

[0069] The model reasoning device of this embodiment is used to implement the substation robot inspection system as described above. Its advantages over the prior art are the same as the advantages of the above-mentioned substation robot inspection system over the prior art, and will not be repeated here.

[0070] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc. In the present application, the unit described as a separate component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment of the present invention. In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0071] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A substation robot inspection system, characterized in that: It includes a robot body, a control device and a thermal imaging device, wherein the control device and the thermal imaging device are arranged on the robot body, and the control device is respectively connected to the thermal imaging device and the robot body for communication: The control device is used to: Obtaining an inspection route and inspection plan for the substation, wherein the inspection plan includes target equipment to be inspected and a preset inspection time; Controlling the robot body to inspect the target device along the inspection route according to the inspection time, and when reaching the location of the target device, collecting a thermal infrared image of the target device through the thermal imaging device; Input the thermal infrared image into a trained electrical component detection model, and output position information corresponding to each electrical component in the thermal infrared image; Determine the current temperature data of the electrical component according to the position information corresponding to the electrical component; and determine whether the electrical component is an abnormal electrical component according to a comparison result between the acquired historical temperature data of the electrical component and the current temperature data; An abnormal inspection report of the target device is generated according to all abnormal electrical components in the target device and corresponding current temperature data.

2. The substation robot inspection system according to claim 1, characterized in that: The inspecting the target device along the inspection route includes: When an obstacle is detected on the inspection route, obtaining boundary position information of the obstacle on the inspection route; generating an obstacle avoidance route according to the boundary position information, wherein a starting point and an end point of the obstacle avoidance route are located on the inspection route; The robot body is controlled to travel from the starting point along the obstacle avoidance route to the end point.

3. The substation robot inspection system according to claim 2, characterized in that: The generating an obstacle avoidance route according to the boundary position information comprises: Determine the minimum circumscribed rectangle corresponding to the obstacle according to the boundary position information; For each vertex of the minimum circumscribed rectangle, respectively determine the point on the obstacle closest to the vertex, and determine the eccentric distance from the point on the obstacle closest to the vertex to the center point of the minimum circumscribed rectangle; Taking the vertex as the center of the circle and taking the eccentric distance corresponding to the point on the obstacle closest to the vertex as the radius, a vertex circle corresponding to the vertex is generated; Determine a tangent line between two adjacent vertex circles, and determine an intersection point between the tangent line and the corresponding vertex circle, wherein the tangent line does not intersect the minimum circumscribed rectangle; Connecting all the tangent intersection points in sequence to generate a closed route; The starting point and the end point of the obstacle avoidance route are determined according to the intersection of the closed route and the inspection route, and the shortest route from the starting point to the end point in the closed route is used as the obstacle avoidance route.

4. The substation robot inspection system according to claim 1, characterized in that: The training process of the electrical component detection model includes: Obtaining historical thermal infrared images of each target device in the substation under different working conditions, and marking the electrical components and corresponding location information in each historical thermal infrared image; Optimizing the grayscale value distribution in the historical thermal infrared image to obtain an optimized historical thermal infrared image; Constructing a training data set based on all the optimized historical thermal infrared images with annotations; The preset initial model is trained according to the training data set to obtain the trained electrical component detection model.

5. The substation robot inspection system according to claim 4, characterized in that: The step of optimizing the historical thermal infrared image according to the gray value distribution in the historical thermal infrared image to obtain the optimized historical thermal infrared image comprises: Performing grayscale processing on the historical thermal infrared image to obtain a grayscale value corresponding to each historical pixel point; Determine the number of pixels corresponding to each grayscale value according to the grayscale values ​​of all the historical pixels; According to all the grayscale values ​​and the corresponding number of pixels, determining a replacement grayscale value corresponding to each grayscale value through a grayscale value optimization relationship; The grayscale value of the corresponding historical pixel point is replaced according to the replacement grayscale value to obtain the optimized historical thermal infrared image; The gray value optimization relationship satisfies: ； Among them, I i is the replacement grayscale value corresponding to the i-th grayscale value, H j is the number of pixels corresponding to the jth grayscale value, N is the number of all pixels in the historical thermal infrared image, N min is the number of pixels corresponding to the minimum grayscale value, and i is the number of the first i grayscale values ​​involved in the superposition calculation.

6. The substation robot inspection system according to claim 1, characterized in that: The determining the current temperature data of the electrical component according to the position information corresponding to the electrical component includes: Determine the grayscale value corresponding to each pixel of the component in the thermal infrared image according to the position information of the electrical component; Determine the temperature value corresponding to the pixel point of the component according to the gray value corresponding to the pixel point of the component and the preset gray value and temperature matching relationship; Determining the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution of the electrical component according to the temperature values ​​corresponding to all the pixel points of the component; The current temperature data is generated according to the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution.

7. The substation robot inspection system according to claim 6, characterized in that: The current temperature distribution satisfies: Determining whether the pixel point of the component is a high-temperature pixel point according to a comparison result between the temperature value of the pixel point of the component and a preset temperature threshold; Based on a plurality of preset partition center points, determining the partition center line closest to the high-temperature pixel point as the partition center point of the high-temperature pixel point; Generate a corresponding high-temperature partition according to all the high-temperature pixel points corresponding to the same partition center point; The current temperature distribution of the electrical component is generated according to all the high temperature zones.

8. The substation robot inspection system according to claim 1, characterized in that: The current temperature data includes the current average temperature, the current maximum temperature, the current minimum temperature and the current temperature distribution; the historical temperature data includes the historical average temperature, the historical maximum temperature, the historical minimum temperature and the historical temperature distribution; the judging whether the electrical component is an abnormal electrical component based on the comparison result of the current temperature data and the historical temperature data includes: According to the current average temperature, the current maximum temperature, the current minimum temperature, the historical average temperature, the historical maximum temperature and the historical minimum temperature, the temperature deviation of the electrical component is obtained through a preset deviation relationship; When the temperature deviation is greater than a preset deviation threshold, determining that the temperature of the electrical component is abnormal; Determining whether the high temperature area is expanding according to the comparison result of the current temperature distribution and the historical temperature distribution, and if so, determining that the heating area of ​​the electrical component is abnormal; When the temperature of the electrical component is abnormal and / or the heating area is abnormal, determining that the electrical component is the abnormal electrical component; The deviation relationship satisfies: ； Wherein, P is the temperature deviation, is the current average temperature, is the historical average temperature, T max is the current maximum temperature, L max is the historical maximum temperature, T min The current minimum temperature, L min The historical minimum temperature.

9. The substation robot inspection system according to claim 8, characterized in that: The determining whether the high temperature area is expanding according to the comparison result between the current temperature distribution and the historical temperature distribution includes: Determine the current high temperature partition area according to all current high temperature partitions in the current temperature distribution, and determine the historical high temperature partition area according to all historical high temperature partitions in the historical temperature distribution; When the current high temperature zone area is larger than the historical high temperature zone area, it is determined that the high temperature region of the electrical component is expanding.

10. A control method for a substation robot inspection system, characterized in that: include: Obtaining an inspection route and inspection plan for the substation, wherein the inspection plan includes target equipment to be inspected and a preset inspection time; Controlling the robot body to inspect the target device along the inspection route according to the inspection time, and when reaching the location of the target device, collecting a thermal infrared image of the target device through a thermal imaging device; Input the thermal infrared image into a trained electrical component detection model, and output position information corresponding to each electrical component in the thermal infrared image; Determine the current temperature data of the electrical component according to the position information corresponding to the electrical component; and determine whether the electrical component is an abnormal electrical component according to a comparison result between the acquired historical temperature data of the electrical component and the current temperature data; An abnormal inspection report of the target device is generated according to all abnormal electrical components in the target device and corresponding current temperature data.

Images