Cable monitoring system and method in cable tunnel based on intelligent identification

Abstract

The present invention discloses a cable monitoring system and method in a cable tunnel based on intelligent identification, belonging to the field of robot technology. The present invention comprises the following steps: S1, obtaining the relationship between the power and the running distance of the robot at different running speeds, and dividing obstacles; S2, establishing a map model, marking points, placing the robot, positioning the robot and updating the map model; S3, screening out the best charging point; S4, screening out the best inspection path; S5, robot inspection, monitoring the temperature rise points on the cable surface; the present invention plans the best inspection path in advance through a comprehensive analysis of the robot and the cable tunnel, and increases the monitoring of the local temperature rise of the cable, which not only takes into account the power of the robot to prevent the robot from running out of power during the inspection process, thereby improving the efficiency of the robot inspection, but also increases the monitoring of the local temperature rise, so that the inspection using the robot is more comprehensive and more practical.

Description

Technical Field

[0001] The present invention relates to the field of robotics technology, and in particular to a cable monitoring system and method in a cable tunnel based on intelligent identification. Background Art

[0002] High-voltage cable tunnels are important urban infrastructure in the power grid. During their operation and maintenance, it is necessary to monitor various comprehensive environmental indicators including cable temperature and partial discharge. The related monitoring work has problems such as harsh working environment, insufficient manpower and detection technology capabilities, and monitoring cost-effectiveness. At present, more and more cable tunnels use quadruped robots for cable inspection.

[0003] However, due to the complex terrain of the cable tunnel, the robot cannot make effective path planning in advance, and lacks consideration of the robot's energy consumption when planning the path. At the same time, during the inspection process, the robot lacks monitoring of the local temperature rise of the cable, which not only affects the robot's inspection efficiency, but also does not fully monitor the cable, reducing the practicality of the robot's inspection.

[0004] Therefore, people are in urgent need of a cable monitoring system in cable tunnels based on intelligent identification to solve the above problems. Summary of the invention

[0005] The object of the present invention is to provide a cable monitoring system in a cable tunnel based on intelligent identification to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] A cable monitoring method in a cable tunnel based on intelligent identification, the method comprising the following steps:

[0008] S1. Collect the running distance of the robot at different running speeds with the same power consumption, and analyze the relationship between the power consumption and running distance of the robot at different running speeds; analyze the obstacle data of the tunnel to be inspected and the minimum travel length of the robot, and divide each obstacle into passable obstacles and impassable obstacles;

[0009] S2. Establish a map model based on the line data of the tunnel to be inspected. In the map model, mark the charging position of the robot as the charging point, mark the position of the obstacle as the obstacle point, and mark the key monitoring position in the tunnel to be inspected as the inspection point. The key monitoring position includes the intersection of the cable line and the position of the power equipment. Place the robot in the tunnel to be inspected, locate the current position of the robot through the sensor carried by the robot combined with the SLAM algorithm, and update the map model of the robot's position in real time.

[0010] S3. Use the Dijkstra algorithm combined with the map model to plan the shortest path from the current position of the robot to each charging point, and obtain the actual length of each shortest path; collect the current power of the robot and obtain the predicted running distance of the robot under this power; select the best charging point according to the length of each shortest path and the predicted running distance;

[0011] S4. Use the breadth-first search algorithm combined with the map model to plan all the paths of the robot from the current position to the best charging point to form a path set, recorded as the first path set; filter the first path set according to the obstacle points to filter out the paths that the robot can pass through to form a new path set, recorded as the second path set; filter the second path set according to the predicted running distance to filter out the paths that meet the predicted running distance to form a new path set, recorded as the third path set; filter the third path set according to the inspection points to filter out the best inspection path;

[0012] S5. The robot conducts inspections along the best inspection route. During the inspection, the dual-spectrum pan-tilt head is used to collect infrared images of the cables. After preprocessing the infrared images, the temperature rise points on the cable surface in the image are obtained through the trained semantic segmentation model. The robot detects the position of the abnormal temperature rise point, marks it as an abnormal point in the map model, and issues an alarm.

[0013] According to the above technical solution, the specific process of step S1 is as follows:

[0014] S1-1. Set the power, recorded as E1, and collect the running distance of the robot at different running speeds that consumes the power E1; according to the following formula, obtain the relationship between the power and running distance of the robot at different running speeds:

[0015]

[0016] Among them, Y α It represents the relationship between the power and running distance of the robot at the αth running speed; S1 α represents the running distance of the robot at the αth running speed, which consumes the power E1; α∈[1, N1], N1 is a positive integer;

[0017] S1-2. Collect obstacle data of the tunnel to be inspected, wherein the obstacle data includes the location of the obstacle and the maximum free length of the location where the obstacle is located. The maximum free length of the location where the βth obstacle is located is recorded as L1. β , β∈[1, N2], N2 is a positive integer;

[0018] S1-3, the minimum travel length of the collection robot, denoted as l;

[0019] S1-4, if L1 is satisfiedβ >l, the obstacle is classified as a passable obstacle; if L1 is not satisfied β >l, the obstacle is classified as an impassable obstacle.

[0020] According to the above technical solution, the specific process of step S2 is as follows:

[0021] S2-1. Build a map model based on the line data of the tunnel to be inspected. Mark the charging position of the robot as the charging point in the map model. Use P1 γ Point represents the γth charging point, γ∈[1, N3], N3 is a positive integer; the location of the obstacle is marked as an obstacle point, and P2 δ Point represents the δth obstacle point, δ∈[1, N4], N4 is a positive integer; the key monitoring position in the tunnel to be inspected is marked as the inspection point, and P3 ε Point represents the εth inspection point, ε∈[1, N5], N5 is a positive integer;

[0022] S2-2, placing the robot in the tunnel to be inspected, recording the placement time as time t, locating the robot at time t through the sensor carried by the robot combined with the SLAM algorithm, recording the robot's position at time t as point O, and updating the map model of the robot's position in real time;

[0023] The SLAM algorithm can collect environmental information by using sensors (such as lidar) and locate the robot's position and the structure of the surrounding environment (update the map) in real time during the robot's movement.

[0024] According to the above technical solution, in step S3, the best charging point is selected according to the length of each shortest path and the predicted running distance. The specific process is as follows:

[0025] S3-1. Use Dijkstra algorithm combined with map model to plan the robot from point O to P1 γ The shortest path from point O to point P1 is obtained according to the scale of the map model. γ The actual length of the shortest path of the point is recorded as

[0026] Dijkstra's algorithm is a classic algorithm for finding the single-source shortest path in a weighted graph. Its main function is to find the shortest path from a starting node to all other nodes in the graph.

[0027] S3-2, the power of the collection robot at time t is recorded as E2, and the predicted running distance of the robot consuming power E2 is as follows:

[0028] S2 α =Yα E2;

[0029] Among them, S2 α The robot's predicted running distance at the αth running speed when consuming power E2;

[0030] According to S2 α , and record the minimum value as S2 min , and the maximum value is recorded as S2 max ;

[0031] S3-3, if satisfy The corresponding charging point is retained; if Dissatisfied Then remove the corresponding charging point;

[0032] The γ'th charging point after screening is represented by P1' γ’ Point, move the robot from point O to P1' γ’ The actual length of the shortest path of the point is recorded as γ'∈[1,N3'], N3' is a positive integer;

[0033] S3-4, use the following formula and S2 α Perform the calculation:

[0034]

[0035] in, Indicates the fit between the γ'th charging point and the predicted running distance, and selects The charging point corresponding to the minimum value of is the optimal charging point and is recorded as point P.

[0036] According to the above technical solution, the specific process of step S4 is as follows:

[0037] S4-1. Use the breadth-first search algorithm combined with the map model to plan all the paths of the robot from point O to point P to form a path set, which is recorded as the first path set;

[0038] The breadth-first search algorithm is a search algorithm used in graph or tree data structures. Its main function is to start from a specified starting node and expand the search outward layer by layer until the target node is found or the entire graph is traversed;

[0039] S4-2, according to P2 δ The paths in the first path set are screened and the paths that contain P2 are selected. δThe path where the obstacle corresponding to the point is an impassable obstacle is removed, and the path through which the robot can pass is screened out to form a new path set, which is recorded as the second path set;

[0040] S4-3. According to the scale of the map model, the path length in the map is converted into the actual length according to the scale, and the actual length of each path in the second path set is obtained. The actual length of the ∈th path in the second path set is recorded as L2 ∈ ,∈∈[1,N6], N6 is a positive integer;

[0041] If L2 ∈ Satisfy S2 min ≤L2 ∈ ≤S2 max , then L2 ∈ The corresponding path is retained; if L2 ∈ Does not satisfy S2 min ≤L2 ∈ ≤S2 max , then L2 ∈ The corresponding path is removed;

[0042] According to the above process, the second path set is screened to form a new path set, which is recorded as the third path set;

[0043] S4-4, according to P3 ε Filter each path in the third path set;

[0044] The third path set contains P3 ε Point and P3 ε The path with the largest number of points is taken as the first result path;

[0045] If the first result path is the only path, the path is used as the best inspection path; if the first result path is not the only path, the first result path is screened again;

[0046] The first result path contains P2 δ Point and P2 δ The path with the least number of points is taken as the second result path;

[0047] If the second result path is the only path, the path is used as the best inspection path; if the second result path is not the only path, the second result path is screened again;

[0048] According to the scale of the map model, the actual length of each path in the second result path is obtained, and the path with the largest actual length in the second result path is used as the optimal inspection path.

[0049] According to the above technical solution, the specific process of step S5 is as follows:

[0050] The robot conducts inspections along the best inspection route. During the inspection process, the dual-spectrum PTZ is used to collect infrared images of the cables. After preprocessing the infrared images, the temperature rise points on the cable surface in the images are obtained through the trained semantic segmentation model. When the robot detects abnormal temperature rise points on the cable surface during the inspection process, the robot's location at that moment is marked as an abnormal point in the map model, and an alarm is issued.

[0051] Multi-scale feature fusion is achieved through image input, feature extraction and convolutional neural network. The semantic segmentation model combines multi-scale features to perform segmentation, thereby improving the recognition ability of objects of different scales. Multi-scale feature fusion helps to better understand the characteristics of the cable surface area and improve the accuracy and robustness of detection.

[0052] A cable monitoring system in a cable tunnel based on intelligent identification, the system comprising an information acquisition module, an information processing module and an information display module;

[0053] The information acquisition module is used to acquire various information required by the system; the information processing module is used to store, analyze and transmit various information of each module; and the information display module is used to display various information of the information processing module.

[0054] According to the above technical solution, the information collection module includes a tunnel information unit and a robot information unit;

[0055] The tunnel information unit is used to collect various data information of the tunnel to be inspected, and the robot information unit is used to collect various data information of the robot.

[0056] According to the above technical solution, the information processing module includes an information storage unit, an information analysis unit and an information transmission unit;

[0057] The information storage unit is used to store various information acquired by the information acquisition module; the information analysis unit is used to analyze various information acquired by the information acquisition module; and the information transmission unit is used for information transmission among various modules in the system.

[0058] According to the above technical solution, the information display module includes a remote display unit and an alarm unit;

[0059] The remote display unit is used to display various information during the robot inspection process, and the alarm unit is used to display alarm information.

[0060] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0061] The present invention conducts a comprehensive analysis of the robot's power, obstacles in the cable tunnel, key monitoring points and the robot's charging position, plans the best inspection route in advance, and adds monitoring of local temperature rise of the cable. It not only takes into account the robot's power, but also prevents the robot from running out of power during the inspection process, thereby improving the efficiency of the robot's inspection, and increases the comprehensiveness and practicality of the robot's inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0063] Figure 1 It is a flow chart of a cable monitoring method in a cable tunnel based on intelligent identification of the present invention;

[0064] Figure 2 It is a structural schematic diagram of a cable monitoring system in a cable tunnel based on intelligent identification of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] See also Figure 1 , the present invention provides a technical solution:

[0067] A cable monitoring method in a cable tunnel based on intelligent identification, the method comprising the following steps:

[0068] S1. Collect the running distance of the robot at different running speeds with the same power consumption, and analyze the relationship between the power consumption and running distance of the robot at different running speeds; analyze the obstacle data of the tunnel to be inspected and the minimum travel length of the robot, and divide each obstacle into passable obstacles and impassable obstacles;

[0069] The specific process of step S1 is as follows:

[0070] S1-1. Set the power, recorded as E1, and collect the running distance of the robot at different running speeds that consumes the power E1; according to the following formula, obtain the relationship between the power and running distance of the robot at different running speeds:

[0071]

[0072] Among them, Y α It represents the relationship between the power and running distance of the robot at the αth running speed; S1 α represents the running distance of the robot at the αth running speed, which consumes the power E1; α∈[1, N1], N1 is a positive integer;

[0073] Obtain the relationship between the robot's power and running distance at different running speeds to provide data support for subsequent path screening;

[0074] S1-2. Collect obstacle data of the tunnel to be inspected, wherein the obstacle data includes the location of the obstacle and the maximum free length of the location where the obstacle is located. The maximum free length of the location where the βth obstacle is located is recorded as L1. β , β∈[1, N2], N2 is a positive integer;

[0075] S1-3, the minimum travel length of the collection robot, denoted as l;

[0076] S1-4, if L1 is satisfied β >l, the obstacle is classified as a passable obstacle; if L1 is not satisfied β >l, the obstacle is classified as an impassable obstacle;

[0077] Obstacles can be divided into different categories, and when planning routes, paths containing obstacles that the robot cannot pass through can be excluded in advance to improve inspection efficiency.

[0078] S2. Establish a map model based on the line data of the tunnel to be inspected. In the map model, mark the charging position of the robot as the charging point, mark the position of the obstacle as the obstacle point, and mark the key monitoring position in the tunnel to be inspected as the inspection point. The key monitoring position includes the intersection of the cable line and the position of the power equipment. Place the robot in the tunnel to be inspected, locate the current position of the robot through the sensor carried by the robot combined with the SLAM algorithm, and update the map model of the robot's position in real time.

[0079] The specific process of step S2 is as follows:

[0080] S2-1. Build a map model based on the line data of the tunnel to be inspected. Mark the charging position of the robot as the charging point in the map model. Use P1 γ Point represents the γth charging point, γ∈[1, N3], N3 is a positive integer; the location of the obstacle is marked as an obstacle point, and P2 δ Point represents the δth obstacle point, δ∈[1, N4], N4 is a positive integer; the key monitoring position in the tunnel to be inspected is marked as the inspection point, and P3 εPoint represents the εth inspection point, ε∈[1, N5], N5 is a positive integer;

[0081] S2-2, placing the robot in the tunnel to be inspected, recording the placement time as time t, locating the robot at time t through the sensor carried by the robot combined with the SLAM algorithm, recording the robot's position at time t as point O, and updating the map model of the robot's position in real time;

[0082] The SLAM algorithm can collect environmental information by using sensors (such as lidar) and locate the robot's position and the structure of the surrounding environment (update the map) in real time during the robot's movement.

[0083] S3. Use the Dijkstra algorithm combined with the map model to plan the shortest path from the current position of the robot to each charging point, and obtain the actual length of each shortest path; collect the current power of the robot and obtain the predicted running distance of the robot under this power; select the best charging point according to the length of each shortest path and the predicted running distance;

[0084] In step S3, the best charging point is selected according to the length of each shortest path and the predicted running distance. The specific process is as follows:

[0085] S3-1. Use Dijkstra algorithm combined with map model to plan the robot from point O to P1 γ The shortest path from point O to point P1 is obtained according to the scale of the map model. γ The actual length of the shortest path of the point is recorded as

[0086] Dijkstra's algorithm is a classic algorithm for finding the single-source shortest path in a weighted graph. Its main function is to find the shortest path from a starting node to all other nodes in the graph.

[0087] S3-2, the power of the collection robot at time t is recorded as E2, and the predicted running distance of the robot consuming power E2 is as follows:

[0088] S2 α =Y α E2;

[0089] Among them, S2 α The robot's predicted running distance at the αth running speed when consuming power E2;

[0090] According to S2 α , and record the minimum value as S2 min , and the maximum value is recorded as S2 max ;

[0091] S3-3, if satisfy The corresponding charging point is retained; if Dissatisfied Then remove the corresponding charging point;

[0092] The γ'th charging point after screening is represented by P1' γ’ Point, move the robot from point O to P1' γ’ The actual length of the shortest path of the point is recorded as γ'∈[1,N3'], N3' is a positive integer;

[0093] S3-4, use the following formula and S2 α Perform the calculation:

[0094]

[0095] in, represents the degree of fit between the γ'th charging point and the predicted running distance;

[0096] Filter out The charging point corresponding to the minimum value of is the optimal charging point and is recorded as point P.

[0097] The process of selecting the best charging point takes not only the battery level but also the inspection distance into consideration, and path planning is performed for the best charging point, which not only enables the robot to complete the inspection according to the inspection route, but also reduces the situation where there is too much remaining power and the inspection distance is too short when the robot's inspection route is completed. It effectively manages the robot's charging, extends the battery life, and at the same time improves the efficiency of the inspection.

[0098] S4. Use the breadth-first search algorithm combined with the map model to plan all the paths of the robot from the current position to the best charging point to form a path set, recorded as the first path set; filter the first path set according to the obstacle points to filter out the paths that the robot can pass through to form a new path set, recorded as the second path set; filter the second path set according to the predicted running distance to filter out the paths that meet the predicted running distance to form a new path set, recorded as the third path set; filter the third path set according to the inspection points to filter out the best inspection path;

[0099] The specific process of step S4 is as follows:

[0100] S4-1. Use the breadth-first search algorithm combined with the map model to plan all the paths of the robot from point O to point P to form a path set, which is recorded as the first path set;

[0101] The breadth-first search algorithm is a search algorithm used in graph or tree data structures. Its main function is to start from a specified starting node and expand the search outward layer by layer until the target node is found or the entire graph is traversed;

[0102] S4-2, according to P2 δ The paths in the first path set are screened and the paths that contain P2 are selected. δ The path where the obstacle corresponding to the point is an impassable obstacle is removed, and the path through which the robot can pass is screened out to form a new path set, which is recorded as the second path set;

[0103] S4-3. According to the scale of the map model, the path length in the map is converted into the actual length according to the scale, and the actual length of each path in the second path set is obtained. The actual length of the ∈th path in the second path set is recorded as L2 ∈ ,∈∈[1,N6], N6 is a positive integer;

[0104] If L2 ∈ Satisfy S2 min ≤L2 ∈ ≤S2 max , then L2 ∈ The corresponding path is retained; if L2 ∈ Does not satisfy S2 min ≤L2 ∈ ≤S2 max , then L2 ∈ The corresponding path is removed;

[0105] According to the above process, the second path set is screened to form a new path set, which is recorded as the third path set;

[0106] S4-4, according to P3 ε Filter each path in the third path set;

[0107] The third path set contains P3 ε Point and P3 ε The path with the largest number of points is taken as the first result path;

[0108] If the first result path is the only path, the path is used as the best inspection path; if the first result path is not the only path, the first result path is screened again;

[0109] The first result path contains P2 δ Point and P2 δ The path with the least number of points is taken as the second result path;

[0110] If the second result path is the only path, the path is used as the best inspection path; if the second result path is not the only path, the second result path is screened again;

[0111] According to the scale of the map model, the actual length of each path in the second result path is obtained, and the path with the largest actual length in the second result path is used as the optimal inspection path.

[0112] S5. The robot conducts inspections along the best inspection route. During the inspection, the dual-spectrum PTZ is used to collect infrared images of the cables. After preprocessing the infrared images, the temperature rise points on the cable surface in the images are obtained through the trained semantic segmentation model. The robot detects the abnormal temperature rise points and marks them as abnormal points in the map model, and issues an alarm.

[0113] The specific process of step S5 is as follows:

[0114] The robot conducts inspections along the best inspection route. During the inspection process, the dual-spectrum PTZ is used to collect infrared images of the cables. After preprocessing the infrared images, the temperature rise points on the cable surface in the images are obtained through the trained semantic segmentation model. When the robot detects abnormal temperature rise points on the cable surface during the inspection process, the robot's location at that moment is marked as an abnormal point in the map model, and an alarm is issued.

[0115] Multi-scale feature fusion is achieved through image input, feature extraction and convolutional neural network. The semantic segmentation model combines multi-scale features to perform segmentation, thereby improving the recognition ability of objects of different scales. Multi-scale feature fusion helps to better understand the characteristics of the cable surface area and improve the accuracy and robustness of detection.

[0116] See also Figure 2 , a cable monitoring system in a cable tunnel based on intelligent identification, the system includes an information acquisition module, an information processing module and an information display module;

[0117] The information collection module is used to collect various information required by the system;

[0118] The information collection module includes a tunnel information unit and a robot information unit;

[0119] The tunnel information unit is used to collect various data information of the tunnel to be inspected, and the robot information unit is used to collect various data information of the robot.

[0120] The information processing module is used to store, analyze and transmit various information of each module;

[0121] The information processing module includes an information storage unit, an information analysis unit and an information transmission unit;

[0122] The information storage unit is used to store various information acquired by the information acquisition module; the information analysis unit is used to analyze various information acquired by the information acquisition module; and the information transmission unit is used for information transmission among various modules in the system.

[0123] The information display module is used to display various information of the information processing module;

[0124] The information display module includes a remote display unit and an alarm unit;

[0125] The remote display unit is used to display various information during the robot inspection process, and the alarm unit is used to display alarm information.

[0126] Example:

[0127] It is known that the Y α Including Y1=20 and Y2=30, use Dijkstra algorithm combined with map model to plan the robot from point O to P1 γ The shortest path of the point, the P1 γ Points include P11 and P12. According to the scale of the map model, get the robot from point O to P1 γ The actual length of the shortest path of the point is recorded as Said include and

[0128] The power of the collection robot at time t is recorded as E2 = 60, and the predicted running distance of the robot consuming power E2 is as follows:

[0129] S2 α =Y α E2;

[0130] S21=1200 and S22=1800 are obtained;

[0131] According to S2 α , and record the minimum value as S2 min =1200, and the maximum value is recorded as S2 max =1800;

[0132] satisfy Keep point P11; satisfy Keep point P12; Dissatisfied Remove point P13;

[0133] P1' after screening γ’Points include P1'1 and P1'2, corresponding to include and

[0134] Use the following formula to and S2 α Perform the calculation:

[0135]

[0136] Get and Filter out The charging point corresponding to the minimum value is point P1'2, and point P1'2 is taken as the optimal charging point.

[0137] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0138] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A cable monitoring method in a cable tunnel based on intelligent identification, characterized in that: The method comprises the following steps: S1. Collect the running distance of the robot at different running speeds with the same power consumption, and analyze the relationship between the power consumption and running distance of the robot at different running speeds; analyze the obstacle data of the tunnel to be inspected and the minimum travel length of the robot, and divide each obstacle into passable obstacles and impassable obstacles; S2. A map model is established based on the line data of the tunnel to be inspected. In the map model, the charging position of the robot is marked as a charging point, the position of the obstacle is marked as an obstacle point, and the key monitoring position in the tunnel to be inspected is marked as an inspection point; the robot is placed in the tunnel to be inspected, and the current position of the robot is located by using the sensor carried by the robot in combination with the SLAM algorithm, and the map model of the robot's position is updated in real time; S3. Use the Dijkstra algorithm combined with the map model to plan the shortest path from the current position of the robot to each charging point, and obtain the actual length of each shortest path; collect the current power of the robot and obtain the predicted running distance of the robot under this power; select the best charging point according to the length of each shortest path and the predicted running distance; S4. Use the breadth-first search algorithm combined with the map model to plan all the paths of the robot from the current position to the best charging point to form a path set, recorded as the first path set; filter the first path set according to the obstacle points to filter out the paths that the robot can pass through to form a new path set, recorded as the second path set; filter the second path set according to the predicted running distance to filter out the paths that meet the predicted running distance to form a new path set, recorded as the third path set; filter the third path set according to the inspection points to filter out the best inspection path; S5. The robot conducts inspections along the best inspection route. During the inspection, the dual-spectrum pan-tilt head is used to collect infrared images of the cables. After preprocessing the infrared images, the temperature rise points on the cable surface in the image are obtained through the trained semantic segmentation model. The robot detects the position of the abnormal temperature rise point, marks it as an abnormal point in the map model, and issues an alarm.

2. The cable monitoring method in a cable tunnel based on intelligent identification according to claim 1 is characterized in that: The specific process of step S1 is as follows: S1-1. Set the power, recorded as E1, and collect the running distance of the robot at different running speeds that consumes the power E1; according to the following formula, obtain the relationship between the power and running distance of the robot at different running speeds: Among them, Y α It represents the relationship between the power and running distance of the robot at the αth running speed; S1 α represents the running distance of the robot at the αth running speed, which consumes the power E1; α∈[1, N1], N1 is a positive integer; S1-2. Collect obstacle data of the tunnel to be inspected, wherein the obstacle data includes the location of the obstacle and the maximum free length of the location where the obstacle is located. The maximum free length of the location where the βth obstacle is located is recorded as L1. β , β∈[1, N2], N2 is a positive integer; S1-3, the minimum travel length of the collection robot, denoted as l; S1-4, if L1 is satisfied β >l, the obstacle is classified as a passable obstacle; if L1 is not satisfied β >l, the obstacle is classified as an impassable obstacle.

3. The cable monitoring method in a cable tunnel based on intelligent identification according to claim 2 is characterized in that: The specific process of step S2 is as follows: S2-1. Build a map model based on the line data of the tunnel to be inspected. Mark the charging position of the robot as the charging point in the map model. Use P1 γ Point represents the γth charging point, γ∈[1, N3], N3 is a positive integer; the location of the obstacle is marked as an obstacle point, and P2 δ Point represents the δth obstacle point, δ∈[1,N4], N4 is a positive integer; Mark the key monitoring locations in the tunnel to be inspected as inspection points and use P3 ε Point represents the εth inspection point, ε∈[1, N5], N5 is a positive integer; S2-2. Place the robot in the tunnel to be inspected, and record the placement time as time t. Use the robot's onboard sensors combined with the SLAM algorithm to locate the robot's position at time t, record the robot's position at time t as point O, and update the map model of the robot's position in real time.

4. The cable monitoring method in a cable tunnel based on intelligent identification according to claim 3 is characterized in that: In step S3, the best charging point is selected according to the length of each shortest path and the predicted running distance. The specific process is as follows: S3-1. Use Dijkstra algorithm combined with map model to plan the robot from point O to P1 γ The shortest path from point O to point P1 is obtained according to the scale of the map model. γ The actual length of the shortest path of the point is recorded as S3-2, the power of the collection robot at time t is recorded as E2, and the predicted running distance of the robot consuming power E2 is as follows: <h2 style=";text-align:left;direction:ltr">S2<h2 style=";text-align:left;direction:ltr"> α <h2 style=";text-align:left;direction:ltr"> =Y<h2 style=";text-align:left;direction:ltr"> α <h2 style=";text-align:left;direction:ltr"> E2; Among them, S2 α The robot's predicted running distance at the αth running speed when consuming power E2; According to S2 α , and record the minimum value as S2 min , and the maximum value is recorded as S2 max ; S3-3, if satisfy The corresponding charging point is retained; if Dissatisfied Then remove the corresponding charging point; The γ'th charging point after screening is represented by P1' γ’ Point, move the robot from point O to P1' γ’ The actual length of the shortest path of the point is recorded as γ'∈[1,N3'], N3' is a positive integer; S3-4, use the following formula and S2 α Perform the calculation: in, represents the degree of fit between the γ'th charging point and the predicted running distance; Filter out The charging point corresponding to the minimum value of is the optimal charging point and is recorded as point P.

5. The cable monitoring method in a cable tunnel based on intelligent identification according to claim 4 is characterized in that: The specific process of step S4 is as follows: S4-1. Use the breadth-first search algorithm combined with the map model to plan all the paths of the robot from point O to point P to form a path set, which is recorded as the first path set; S4-2, according to P2 δ The paths in the first path set are screened and the paths that contain P2 are selected. δ The path where the obstacle corresponding to the point is an impassable obstacle is removed, and the path through which the robot can pass is screened out to form a new path set, which is recorded as the second path set; S4-3. According to the scale of the map model, obtain the actual length of each path in the second path set, and record the actual length of the ∈th path in the second path set as L2 ∈ ,∈∈[1,N6], N6 is a positive integer; If L2 ∈ Satisfy S2 min ≤L2 ∈ ≤S2 max , then L2 ∈ The corresponding path is retained; if L2 ∈ Does not satisfy S2 min ≤L2 ∈ ≤S2 max , then L2 ∈ The corresponding path is removed; According to the above process, the second path set is screened to form a new path set, which is recorded as the third path set; S4-4, according to P3 ε Filter each path in the third path set; The third path set contains P3 ε Point and P3 ε The path with the largest number of points is taken as the first result path; If the first result path is the only path, then the path is used as the best inspection path; If the first result path is not the only path, the first result path is screened again; The first result path contains P2 δ Point and P2 δ The path with the least number of points is taken as the second result path; If the second result path is the only path, then the path is used as the best inspection path; If the second result path is not the only path, the second result path is screened again; According to the scale of the map model, the actual length of each path in the second result path is obtained, and the path with the largest actual length in the second result path is used as the optimal inspection path.

6. The cable monitoring method in a cable tunnel based on intelligent identification according to claim 5 is characterized in that: The specific process of step S5 is as follows: The robot conducts inspections along the optimal inspection route. During the inspection process, a dual-spectrum pan-tilt head is used to collect infrared images of the cables. After preprocessing the infrared images, the temperature rise points on the cable surface in the image are obtained through a trained semantic segmentation model. When the robot detects abnormal temperature rise points on the cable surface during the inspection, the robot's location at that moment is marked as an abnormal point in the map model, and an alarm is issued.

7. A cable monitoring system in a cable tunnel based on intelligent identification, such as using the cable monitoring method in a cable tunnel based on intelligent identification according to any one of claims 1 to 5, characterized in that: The system includes an information acquisition module, an information processing module and an information display module; The information acquisition module is used to acquire various information required by the system; the information processing module is used to store, analyze and transmit various information of each module; and the information display module is used to display various information of the information processing module.

8. The cable monitoring system in a cable tunnel based on intelligent identification according to claim 7 is characterized in that: The information collection module includes a tunnel information unit and a robot information unit; The tunnel information unit is used to collect various data information of the tunnel to be inspected, and the robot information unit is used to collect various data information of the robot.

9. The cable monitoring system in a cable tunnel based on intelligent identification according to claim 7, characterized in that: The information processing module includes an information storage unit, an information analysis unit and an information transmission unit; The information storage unit is used to store various information acquired by the information acquisition module; the information analysis unit is used to analyze various information acquired by the information acquisition module; and the information transmission unit is used for information transmission among various modules in the system.

10. The cable monitoring system in a cable tunnel based on intelligent identification according to claim 7, characterized in that: The information display module includes a remote display unit and an alarm unit; The remote display unit is used to display various information during the robot inspection process, and the alarm unit is used to display alarm information.

Images