| In recent years,with the rapid development of China’s rail transit industry,the safety and reliable operation of trains is the top priority,which is directly related to the life,health and safety of passengers.Most of the key components of train are distributed under the train carriage.Regular inspection and maintenance are the main means to prevent component failure and the important guarantees to ensure the stable and reliable operation of trains.The existing inspection of train underbody components mainly depends on manual work.The working environment is difficult,and there are many kinds of underbody components.Human negligence may easily lead to missed inspection and false inspection,which lays a hidden danger for the safety of train operation.This thesis takes the six-degree-of-freedom manipulator based on ROS carried by the train intelligent inspection robot as the research object.The main work of this thesis is as follows:Firstly,the pose of UR5 e manipulator is analyzed,the standard D-H parameters of the manipulator are solved,the kinematics analysis simulation model of the manipulator is established,the transformation matrix and joint angle value of the forward and inverse kinematics of the manipulator are analyzed and programmed,the accuracy of the forward and inverse kinematics analysis results of UR5 e manipulator is verified by Robotics Tools,and the reachable workspace of the manipulator is verified by multiple random points.Subsequently,in order to accurately obtain the influence of the train intelligent inspection robot parking error on the planned path of the manipulator,the 3D point cloud registration algorithm is studied,in view of the low efficiency of 3D point cloud registration based on ICP algorithm,an improved ICP algorithm is designed to register the 3D point cloud data after rotation translation transformation with the origin cloud data of the model,the 3D point cloud data exceeding the point cloud deviation threshold is voxelized,and the point cloud data of hypothetical obstacles are modeled by appropriate bounding box algorithm.The structure of the manipulator is simplified,the collision detection problem between the manipulator and the external environment is abstracted as the axial intersection problem,and the obstacle avoidance ability of the manipulator is tested.Then,combined with the advantages and disadvantages of RRT algorithm and some improved RRT algorithms,an improved path planning algorithm based on RRT algorithm and curve fitting,namely VO-RRT algorithm,is designed by introducing variable step size optimization strategy,point oriented expansion strategy,initial path optimization strategy and path smoothing strategy.The path nodes solution experiments of 2D and 3D obstacle space are carried out in MATLAB,and the improved performance of VO-RRT algorithm is compared and tested,the path nodes obtained by VO-RRT algorithm has enhanced obstacle avoidance ability and smoothness in different obstacle spaces,the simulation test of manipulator path planning is carried out in ROS,and the feasibility of the path planned by VO-RRT algorithm in UR5 e manipulator is verified.Finally,according to the environmental characteristics of the station track,the inspection scheme planning of the manipulator for train underbody components is completed,the overall function of the train intelligent inspection robot is analyzed and designed,the model of the train intelligent inspection robot and the simplified model of its working environment are built in the 3D physical simulation platform of ROS,the emergency remote control functions of the automated guided vehicle and the manipulator are designed,in the simulation environment,the flow of the train intelligent inspection robot to inspect train underbody components is tested,and the feasibility of the inspection strategy and path planning effect is verified. |