Inspection robot dynamic obstacle avoidance planning system and method based on three-dimensional point cloud

Abstract

The invention discloses an inspection robot dynamic obstacle avoidance planning system and method based on a three-dimensional point cloud, relates to the technical field of dynamic obstacle avoidance planning, and solves the technical problems that in the prior art, trajectory prediction cannot be performed on an obstacle after the obstacle is recognized, and targeted obstacle avoidance cannot be performed in combination with obstacle prediction. Specifically, the obstacle recognition module performs obstacle recognition on the inspection process of the inspection robot; the obstacle trajectory prediction unit is used for performing trajectory prediction on obstacles on the inspection trajectory; and the obstacle avoidance decision analysis unit performs obstacle avoidance decision analysis on the inspection robot.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic obstacle avoidance planning, and in particular to a dynamic obstacle avoidance planning system and method for an inspection robot based on three-dimensional point cloud. Background Art

[0002] A three-dimensional point cloud is a dataset consisting of a large number of discrete points in space. Each point contains three-dimensional coordinates (X, Y, Z), as well as possible attributes such as color, reflection intensity, texture, and timestamp. The point cloud obtained by lidar scanning of buildings can restore the three-dimensional outlines of walls and windows. Dynamic obstacle avoidance refers to the process of a robot sensing the environment through sensors and adjusting its path in real time to deal with real-time changing obstacles (such as moving people, vehicles, other robots, etc.) during movement.

[0003] However, in the existing technology, it is impossible to accurately identify obstacles, and it is impossible to predict the obstacle trajectory after identifying the obstacle, and it is impossible to combine the obstacle prediction for targeted obstacle avoidance. In addition, after obtaining the predicted movement trajectory of the obstacle, it is impossible to make targeted obstacle avoidance decisions in combination with the inspection task of the inspection robot.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to solve the above-mentioned problems and to propose a dynamic obstacle avoidance planning system and method for an inspection robot based on three-dimensional point cloud.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A dynamic obstacle avoidance planning system for an inspection robot based on a three-dimensional point cloud includes a dynamic obstacle avoidance planning center, wherein the dynamic obstacle avoidance planning center is communicatively connected to an obstacle recognition module, an obstacle trajectory prediction unit, and an obstacle avoidance decision analysis unit; The obstacle recognition module identifies obstacles during the inspection process of the inspection robot; The obstacle trajectory prediction unit predicts the trajectory of obstacles on the inspection trajectory; The obstacle avoidance decision analysis unit performs obstacle avoidance decision analysis on the inspection robot.

[0007] As a preferred embodiment of the present invention, the process of the obstacle recognition module is as follows: The starting point of the inspection robot is set as the coordinate origin, and a three-dimensional space coordinate system is constructed based on the coordinate origin and the inspection area. The coordinates of objects in the area where the inspection robot is located during the inspection process are marked using three-dimensional point cloud technology, and the coordinates of each mark are analyzed according to the inspection process; Radar detection is performed based on the real-time position of the inspection robot, and radar reflected waves are collected from objects at each point in the spatial coordinate system. When the coordinates of the inspection robot continue to change, the azimuth distance between the coordinate point object inferred by the reflected wave in the corresponding moving trajectory and the inspection robot is obtained. Based on the azimuth distance, the overlapping area between the position of the corresponding object and any point on the moving trajectory during the movement of the inspection robot is obtained. At the same time, the maximum deviation value between the changing speed of the corresponding object position and the moving speed of the inspection robot during the movement of the inspection robot is obtained.

[0008] As a preferred embodiment of the present invention, the collected data is analyzed: If the overlapping area between the position of the corresponding object and any point on the movement trajectory during the movement of the inspection robot does not exceed the overlapping area threshold, and the maximum deviation between the speed of change of the corresponding object's position and the movement speed of the inspection robot during the movement of the inspection robot does not exceed the maximum speed deviation threshold, an obstacle non-marking signal is generated and sent to the dynamic obstacle avoidance planning center together with the object's three-dimensional coordinates; after the inspection trajectory changes, monitoring of this type of object continues; If the overlapping area between the position of the corresponding object and any point on the moving trajectory during the movement of the inspection robot exceeds the overlapping area threshold, or the maximum deviation between the changing speed of the corresponding object and the moving speed of the inspection robot during the movement of the inspection robot exceeds the maximum speed deviation threshold, an obstacle marking signal is generated and sent to the dynamic obstacle avoidance planning center together with the three-dimensional coordinates of the object. After receiving the signal, the dynamic obstacle avoidance planning center generates an obstacle trajectory prediction signal and sends it to the obstacle trajectory prediction unit.

[0009] As a preferred embodiment of the present invention, the process of the obstacle trajectory prediction unit is as follows: Determine the current 3D coordinates of the obstacle and predict its trajectory based on the current position. Assuming that the obstacle moves in a uniform straight line, the position at the future time t+Δt is: , It is expressed as the predicted position of the obstacle at the future time t+Δt; P t It represents the estimated position of the obstacle at the current time t; Vt represents the estimated speed of the obstacle at the current time t; Δt represents the time interval, that is, the time difference from the current time t to the future time t+Δt; During the movement of the obstacle, taking into account the uncertainty, the uncertainty covariance of the position is: ; Expressed as the uncertainty covariance matrix of the position at the future time t+Δt; Expressed as the covariance matrix of the position estimate at the current time t Expressed as the covariance matrix of the velocity estimate.

[0010] As a preferred embodiment of the present invention, the real-time predicted position of the obstacle is marked based on the uncertainty covariance calculation of the position, and the obstacle trajectory is predicted based on the current position of the obstacle and the historical movement trajectory. The predicted trajectory of the obstacle is constructed based on the predicted position and the real-time position combined with the moving speed of the historical trajectory; the real-time predicted trajectory is sent to the obstacle avoidance decision analysis unit.

[0011] As a preferred embodiment of the present invention, the process of the obstacle avoidance decision analysis unit is as follows: Based on the inspection trajectory of the inspection robot, it is inferred whether the location of the obstacle in the current trajectory overlaps with the inspection point. If there is overlap, the obstacle avoidance strategy is set to the staggered strategy; otherwise, if there is no overlap, the obstacle avoidance strategy is set to the track change strategy. Under the staggered strategy, the inspection robot's movement distance increases after obtaining the inspection points that prioritize non-obstacle areas. At the same time, the distance interval between the inspection robot's real-time stop position and the nearest inspection point after it staggers its current trajectory to avoid obstacles is obtained: If the inspection robot prioritizes inspection points that are not in the current obstacle area and the increase in the movement distance of the inspection robot exceeds the distance increase threshold, or if the distance between the real-time stopping position of the inspection robot and the nearest inspection point after the inspection robot deviates from the current trajectory to avoid an obstacle exceeds the distance interval threshold, a speed avoidance signal is generated; If the inspection robot prioritizes inspecting inspection points that are not in the current obstacle area, the increase in the inspection robot's moving distance does not exceed the distance increase span threshold, and the distance interval between the real-time stop position of the inspection robot and the nearest inspection point after it staggers its current trajectory to avoid obstacles does not exceed the distance interval threshold, a staggered strategy execution signal is generated.

[0012] As a preferred embodiment of the present invention, under the track change strategy, the numerical deviation between the track coverage rate in the real-time inspection area and the original track coverage rate after the inspection robot's track is changed is obtained, and at the same time, the cumulative peak value of the overlapping movement distance of the inspection points after the inspection robot's track is changed is obtained: If the numerical deviation between the track coverage rate in the real-time inspection area and the original track coverage rate after the inspection robot’s track is changed exceeds the numerical deviation threshold, or the cumulative peak value of the overlapping movement distance after the inspection points are connected after the inspection robot’s track is changed exceeds the overlapping movement distance threshold, a track change inefficiency signal is generated; If the numerical deviation between the track coverage rate in the real-time inspection area and the original track coverage rate after the inspection robot's track is changed does not exceed the numerical deviation threshold, and the cumulative peak value of the overlapping movement distance after the inspection points are connected after the inspection robot's track is changed does not exceed the overlapping movement distance threshold, a track change efficiency signal is generated.

[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. In the present invention, obstacle recognition is performed during the inspection process of the inspection robot. Through obstacle recognition, data processing can be performed in a timely manner to facilitate obstacle avoidance planning and decision-making, thereby improving the inspection safety of the inspection robot and reflecting the data processing effect of the inspection robot.

[0014] 2. In the present invention, the trajectory of obstacles on the inspection trajectory is predicted, and the obstacle trajectory prediction is used to infer whether the current obstacle can be avoided before the inspection robot arrives. At the same time, when the inspection robot decides to avoid the obstacle, the real-time obstacle avoidance decision trajectory will not conflict with the obstacle trajectory, thereby reducing the obstacle avoidance accuracy of the inspection robot. At the same time, after the obstacle avoidance decision is executed, the inspection robot can efficiently complete the current inspection task.

[0015] 3. In the present invention, obstacle avoidance decision analysis is performed on the inspection robot. By considering the predicted trajectory of the obstacle and combining it with the obstacle avoidance plan of the inspection robot, obstacle avoidance decision is made, which improves the obstacle avoidance efficiency of the inspection robot and can also reduce the risk of obstacles still appearing after obstacle avoidance, so that the inspection operation completion rate of the inspection robot decreases. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0017] Figure 1 This is a system principle block diagram of the present invention; Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0018] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0020] See also Figure 1 As shown, a dynamic obstacle avoidance planning system for an inspection robot based on a three-dimensional point cloud includes a dynamic obstacle avoidance planning center, wherein the dynamic obstacle avoidance planning center is communicatively connected to an obstacle recognition module, an obstacle trajectory prediction unit, and an obstacle avoidance decision analysis unit; The dynamic obstacle avoidance planning center generates obstacle recognition signals and sends them to the obstacle recognition module; After receiving the data, the obstacle recognition module identifies obstacles during the inspection process of the inspection robot. Through obstacle recognition, data processing can be carried out in a timely manner to facilitate obstacle avoidance planning and decision-making, thereby improving the inspection safety of the inspection robot and also demonstrating the data processing effect of the inspection robot; The starting point of the inspection robot is set as the coordinate origin, and a three-dimensional space coordinate system is constructed based on the coordinate origin and the inspection area. The coordinates of objects in the area where the inspection robot is located during the inspection process are marked using three-dimensional point cloud technology, and the coordinates of each mark are analyzed according to the inspection process; The real-time position of the inspection robot is used for radar detection, and radar reflection waves corresponding to objects at various points in the spatial coordinate system are collected. When the coordinates of the inspection robot continue to change, the azimuth distance between the coordinate point object inferred by the reflection wave in the corresponding moving trajectory and the inspection robot is obtained. Based on the azimuth distance, the overlapping area between the position of the corresponding object and any point on the moving trajectory during the movement of the inspection robot is obtained. At the same time, the maximum deviation value between the change speed of the corresponding object position and the movement speed of the inspection robot during the movement of the inspection robot is obtained. And analyze the collected data: If the overlapping area between the position of the corresponding object and any point on the moving trajectory during the movement of the inspection robot does not exceed the overlapping area threshold, and the maximum deviation between the speed of change of the position of the corresponding object and the moving speed of the inspection robot during the movement of the inspection robot does not exceed the maximum speed deviation threshold, it is inferred that the object at the current coordinate system point is not identified as an obstacle on the current inspection trajectory, and an obstacle non-marking signal is generated and sent to the dynamic obstacle avoidance planning center together with the three-dimensional coordinates of the object; after the inspection trajectory changes, the monitoring of this type of object continues; If the overlapping area between the position of the corresponding object and any point on the moving trajectory during the movement of the inspection robot exceeds the overlapping area threshold, or the maximum deviation between the speed of change of the position of the corresponding object and the moving speed of the inspection robot during the movement of the inspection robot exceeds the maximum speed deviation threshold, it is inferred that the object at the current coordinate system point is identified as an obstacle on the current inspection trajectory, and an obstacle marking signal is generated and sent to the dynamic obstacle avoidance planning center together with the three-dimensional coordinates of the object. After receiving the obstacle marking signal, the dynamic obstacle avoidance planning center generates an obstacle trajectory prediction signal and sends it to the obstacle trajectory prediction unit; After receiving the information, the obstacle trajectory prediction unit predicts the trajectory of the obstacle on the inspection trajectory. It infers whether the current obstacle can be avoided before the inspection robot arrives through the obstacle trajectory prediction. At the same time, when the inspection robot decides to avoid the obstacle, the real-time obstacle avoidance decision trajectory will not conflict with the obstacle trajectory, thereby reducing the obstacle avoidance accuracy of the inspection robot. At the same time, after the obstacle avoidance decision is executed, the inspection robot can efficiently complete the current inspection task; Determine the current 3D coordinates of the obstacle and predict its trajectory based on the current position. Assuming that the obstacle moves in a uniform straight line, the position at the future time t+Δt is: , It is expressed as the predicted position of the obstacle at the future time t+Δt; P t It represents the estimated position of the obstacle at the current time t; Vt represents the estimated speed of the obstacle at the current time t; Δt represents the time interval, that is, the time difference from the current time t to the future time t+Δt; During the movement of the obstacle, taking into account the uncertainty, the uncertainty covariance of the position is: ; Expressed as the uncertainty covariance matrix of the position at the future time t+Δt; Expressed as the covariance matrix of the position estimate at the current time t Expressed as the covariance matrix of velocity estimates; The real-time predicted position of the obstacle is marked based on the uncertainty covariance calculation of the position, and the obstacle trajectory is predicted based on the current position and historical movement trajectory of the obstacle. The predicted obstacle trajectory is constructed based on the predicted position and real-time position combined with the movement speed of the historical trajectory; Send the real-time predicted trajectory to the obstacle avoidance decision analysis unit; After receiving the real-time predicted trajectory, the obstacle avoidance decision analysis unit performs obstacle avoidance decision analysis on the inspection robot. By considering the predicted trajectory of the obstacle and combining it with the inspection robot's obstacle avoidance plan, the obstacle avoidance decision is made. This improves the inspection robot's obstacle avoidance efficiency and reduces the risk of obstacles still blocking the inspection robot after obstacle avoidance, which may reduce the inspection completion rate of the inspection robot. Based on the inspection trajectory of the inspection robot, it is inferred whether the location of the obstacle in the current trajectory overlaps with the inspection point. If there is overlap, the obstacle avoidance strategy is set to the staggered strategy; otherwise, if there is no overlap, the obstacle avoidance strategy is set to the track change strategy. Under the staggered strategy, the inspection robot's movement distance increase span is obtained after the inspection robot prioritizes inspection points in areas other than the current obstacle. At the same time, the distance interval between the real-time stop position and the nearest inspection point after the inspection robot staggers its current trajectory to avoid obstacles is obtained. The movement distance increase span of the inspection robot after the inspection robot prioritizes inspection points in areas other than the current obstacle and the distance interval between the real-time stop position and the nearest inspection point after the inspection robot staggers its current trajectory to avoid obstacles are compared with the distance increase span threshold and the distance interval threshold, respectively: If the inspection robot prioritizes inspecting inspection points that are not in the current obstacle area, and the increase in the inspection robot's moving distance exceeds the distance increase span threshold, or if the inspection robot deviates from the current trajectory to avoid an obstacle, the distance interval between the real-time stop position and the nearest inspection point exceeds the distance interval threshold, it is inferred that the current dislocation strategy of the inspection robot cannot be executed, and a speed avoidance signal is generated and sent to the dynamic obstacle avoidance planning center; after receiving the speed avoidance signal, the dynamic obstacle avoidance planning center reduces the inspection robot's moving speed and improves the accuracy of the current trajectory approaching the obstacle, thereby reducing the inspection robot's re-inspection time by reducing the current speed and improving the accuracy; If the inspection robot prioritizes inspection points that are not in the current obstacle area, and the increase in the inspection robot's moving distance does not exceed the distance increase span threshold, and the distance interval between the real-time stop position and the nearest inspection point after the inspection robot staggers its current trajectory to avoid obstacles does not exceed the distance interval threshold, it is inferred that the current inspection robot's stagger strategy can be executed, and a stagger strategy execution signal is generated and sent to the dynamic obstacle avoidance planning center; Under the track change strategy, the numerical deviation between the track coverage rate and the original track coverage rate in the real-time inspection area after the inspection robot's track is changed is obtained, and the cumulative peak value of the overlapping movement distance after the inspection points are connected after the inspection robot's track is changed is obtained. The numerical deviation between the track coverage rate and the original track coverage rate in the real-time inspection area after the inspection robot's track is changed, and the cumulative peak value of the overlapping movement distance after the inspection points are connected after the inspection robot's track is changed are compared with the numerical deviation threshold and the overlapping movement distance threshold respectively: If the numerical deviation between the real-time track coverage rate and the original track coverage rate in the inspection area after the inspection robot's trajectory changes exceeds the numerical deviation threshold, or the cumulative peak of the overlapping movement distance after the inspection points are connected after the inspection robot's trajectory changes exceeds the overlapping movement distance threshold, it is inferred that the inspection robot's track change strategy is inefficient, and a track change inefficiency signal is generated and sent to the dynamic obstacle avoidance planning center; If the numerical deviation between the real-time track coverage rate and the original track coverage rate in the inspection area after the inspection robot's trajectory changes does not exceed the numerical deviation threshold, and the cumulative peak value of the overlapping movement distance after the inspection points are connected after the inspection robot's trajectory changes does not exceed the overlapping movement distance threshold, it is inferred that the inspection robot's track change strategy is executed efficiently, and a track change efficiency signal is generated and sent to the dynamic obstacle avoidance planning center; After receiving the information, the dynamic obstacle avoidance planning center will directly execute the track change if it is efficient. If it is inefficient, it will partially change the track based on the real-time obstacle position, and conduct inspections along the original trajectory after moving away from the influence of the obstacle.

[0021] See also Figure 2 As shown in the figure, the dynamic obstacle avoidance planning method of the inspection robot based on three-dimensional point cloud is as follows: Obstacle recognition: perform obstacle recognition during the inspection process of the inspection robot; Obstacle trajectory prediction unit, which predicts the trajectory of obstacles on the inspection track; Obstacle avoidance decision analysis, perform obstacle avoidance decision analysis on the inspection robot.

[0022] When the present invention is in use, the obstacle recognition module performs obstacle recognition on the inspection process of the inspection robot; the obstacle trajectory prediction unit performs trajectory prediction on obstacles on the inspection trajectory; and the obstacle avoidance decision analysis unit performs obstacle avoidance decision analysis on the inspection robot.

[0023] Thresholds, preset values, and preset ranges are set for comparative analysis of results to determine whether they are good or bad. The values ​​are set based on a combination of large-scale model analysis of sample data and manual experience, and can also be adjusted appropriately based on seasonal or common-sense factors. The settings of weight ratio coefficients, influencing factors, etc. are assigned specific values ​​according to the influence of each parameter on the result, which ultimately reflects the impact on the result. They are also set and entered into storage through a combination of large-scale model analysis of sample data and manual experience. Appropriate adjustments can also be made based on seasonal or common-sense influencing conditions.

[0024] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A dynamic obstacle avoidance planning system for inspection robots based on three-dimensional point clouds, characterized by: It includes a dynamic obstacle avoidance planning center, wherein the dynamic obstacle avoidance planning center is communicatively connected to an obstacle recognition module, an obstacle trajectory prediction unit, and an obstacle avoidance decision analysis unit; The obstacle recognition module identifies obstacles during the inspection process of the inspection robot; The obstacle trajectory prediction unit predicts the trajectory of obstacles on the inspection trajectory; The obstacle avoidance decision analysis unit performs obstacle avoidance decision analysis on the inspection robot.

2. The dynamic obstacle avoidance planning system for inspection robots based on three-dimensional point clouds according to claim 1 is characterized in that: The process of the obstacle recognition module is as follows: The starting point of the inspection robot is set as the coordinate origin, and a three-dimensional space coordinate system is constructed based on the coordinate origin and the inspection area. The coordinates of objects in the area where the inspection robot is located during the inspection process are marked using three-dimensional point cloud technology, and the coordinates of each mark are analyzed according to the inspection process; The inspection robot performs radar detection based on its real-time position and collects radar reflection waves from objects at each point in the spatial coordinate system. When the coordinates of the inspection robot continue to change, the azimuth distance between the coordinate point object inferred by the reflected wave in the corresponding moving trajectory and the inspection robot is obtained. According to the azimuth distance, the overlapping area between the position of the corresponding object and any point on the moving trajectory during the movement of the inspection robot is obtained. At the same time, the maximum deviation value between the changing speed of the corresponding object position and the moving speed of the inspection robot during the movement of the inspection robot is obtained.

3. The dynamic obstacle avoidance planning system for inspection robots based on three-dimensional point clouds according to claim 2 is characterized in that: And analyze the collected data: If the overlapping area between the position of the corresponding object and any point on the movement trajectory during the movement of the inspection robot does not exceed the overlapping area threshold, and the maximum deviation between the speed of change of the corresponding object's position and the movement speed of the inspection robot during the movement of the inspection robot does not exceed the maximum speed deviation threshold, an obstacle non-marking signal is generated and sent to the dynamic obstacle avoidance planning center together with the object's three-dimensional coordinates; after the inspection trajectory changes, monitoring of this type of object continues; If the overlapping area between the position of the corresponding object and any point on the moving trajectory during the movement of the inspection robot exceeds the overlapping area threshold, or the maximum deviation between the changing speed of the corresponding object and the moving speed of the inspection robot during the movement of the inspection robot exceeds the maximum speed deviation threshold, an obstacle marking signal is generated and sent to the dynamic obstacle avoidance planning center together with the three-dimensional coordinates of the object. After receiving the signal, the dynamic obstacle avoidance planning center generates an obstacle trajectory prediction signal and sends it to the obstacle trajectory prediction unit.

4. The dynamic obstacle avoidance planning system for inspection robots based on three-dimensional point clouds according to claim 3 is characterized in that: The process of the obstacle trajectory prediction unit is as follows: Determine the current 3D coordinates of the obstacle and predict its trajectory based on the current position. Assuming that the obstacle moves in a uniform straight line, the position at the future time t+Δt is: , It is expressed as the predicted position of the obstacle at the future time t+Δt; P t It represents the estimated position of the obstacle at the current time t; Vt represents the estimated speed of the obstacle at the current time t; Δt represents the time interval, that is, the time difference from the current time t to the future time t+Δt; During the movement of the obstacle, taking into account the uncertainty, the uncertainty covariance of the position is: ; Expressed as the uncertainty covariance matrix of the position at the future time t+Δt; Expressed as the covariance matrix of the position estimate at the current time t Expressed as the covariance matrix of the velocity estimate.

5. The dynamic obstacle avoidance planning system for inspection robots based on three-dimensional point clouds according to claim 4 is characterized in that: The real-time predicted position of the obstacle is marked based on the uncertainty covariance calculation of the position, and the obstacle's trajectory is predicted based on the obstacle's current position and historical movement trajectory. The predicted obstacle trajectory is constructed based on the predicted position and real-time position combined with the moving speed of the historical trajectory; the real-time predicted trajectory is sent to the obstacle avoidance decision analysis unit.

6. The dynamic obstacle avoidance planning system for inspection robots based on three-dimensional point clouds according to claim 5 is characterized in that: The process of the obstacle avoidance decision analysis unit is as follows: Based on the inspection trajectory of the inspection robot, it is inferred whether the location of the obstacle in the current trajectory overlaps with the inspection point. If there is overlap, the obstacle avoidance strategy is set to the staggered strategy; otherwise, if there is no overlap, the obstacle avoidance strategy is set to the track change strategy. Under the staggered strategy, the inspection robot's movement distance increases after obtaining the inspection points that prioritize non-obstacle areas. At the same time, the distance interval between the inspection robot's real-time stop position and the nearest inspection point after it staggers its current trajectory to avoid obstacles is obtained: If the inspection robot prioritizes inspection points that are not in the current obstacle area and the increase in the movement distance of the inspection robot exceeds the distance increase threshold, or if the distance between the real-time stopping position of the inspection robot and the nearest inspection point after the inspection robot deviates from the current trajectory to avoid an obstacle exceeds the distance interval threshold, a speed avoidance signal is generated; If the inspection robot prioritizes inspecting inspection points that are not in the current obstacle area, the increase in the inspection robot's moving distance does not exceed the distance increase span threshold, and the distance interval between the real-time stop position of the inspection robot and the nearest inspection point after it staggers its current trajectory to avoid obstacles does not exceed the distance interval threshold, a staggered strategy execution signal is generated.

7. The dynamic obstacle avoidance planning system for inspection robots based on three-dimensional point clouds according to claim 6, characterized in that: Under the track change strategy, obtain the numerical deviation between the real-time track coverage rate and the original track coverage rate in the inspection area after the inspection robot's track changes, and at the same time obtain the cumulative peak value of the overlapping movement distance of the inspection points after the inspection robot's track changes: If the numerical deviation between the track coverage rate in the real-time inspection area and the original track coverage rate after the inspection robot’s track is changed exceeds the numerical deviation threshold, or the cumulative peak value of the overlapping movement distance after the inspection points are connected after the inspection robot’s track is changed exceeds the overlapping movement distance threshold, a track change inefficiency signal is generated; If the numerical deviation between the track coverage rate in the real-time inspection area and the original track coverage rate after the inspection robot's track is changed does not exceed the numerical deviation threshold, and the cumulative peak value of the overlapping movement distance after the inspection points are connected after the inspection robot's track is changed does not exceed the overlapping movement distance threshold, a track change efficiency signal is generated.

8. A dynamic obstacle avoidance planning method for inspection robots based on three-dimensional point clouds, characterized in that: The specific dynamic obstacle avoidance planning method is as follows: Obstacle recognition: perform obstacle recognition during the inspection process of the inspection robot; Obstacle trajectory prediction unit, which predicts the trajectory of obstacles on the inspection track; Obstacle avoidance decision analysis, perform obstacle avoidance decision analysis on the inspection robot.

Images