Study On Path Planning Method Of Train Inspection Robot

Train maintenance is a vital part of rail vehicle operation and maintenance.Using robots to replace manual train inspections can effectively solve the problems of missed inspections and false inspections caused by manual train inspections.However,the complexity of the train bottom inspection environment seriously restricts the application and development of the train inspection robot,so it is of great significance to study the motion path planning and collision detection of the train bottom inspection robot in the complex environment.The path planning and collision detection of the train inspection robot mainly involve the modeling of the train inspection environment,the kinematics analysis of the robot,the spatial path planning and collision detection analysis.In this paper,the path planning and collision detection of the 6-DOF train inspection robot in complex environment are studied.Firstly,the train inspection object and train inspection robot in train inspection scene were modeled to determine the work space range of train inspection robot.Combined with the space environment at the bottom of the train,the shape and motion characteristics of the robot manipulator,the kinematics analysis of the train inspection robot was carried out based on the D-H method,and the correctness of the D-H model was verified by MATLAB.After modeling different polynomial interpolation methods,MATLAB was used to compare different levels of polynomial interpolation methods.It was verified that the position curve,velocity curve and acceleration curve of the quintic polynomial interpolation method planning were more stable.Using polynomial interpolation method to plan the space path of the train inspection robot.Secondly,in view of the complex structure of the bottom of the train and the train inspection robot in the real scene,a collision detection method for complex scenes based on dynamic viewpoints is proposed.The existing collision detection algorithms are compared,the AABB bounding box is used for collision detection and analysis,combined with the characteristics of the space segmentation method,the AABB bounding box is spatially segmented using a uniform octree.In view of the fact that traditional collision detection methods cannot meet the requirements of collision detection in real scenes,they will have problems such as low efficiency,large calculation amount,and slow update rate.Taking dynamic objects as the viewpoint,a collision detection method based on dynamic viewpoints is proposed.The use of local collision detection instead of global collision detection effectively solves the construction of unnecessary models in the collision detection process and improves the efficiency of collision detection.Finally,VS2015 and OSG were used to build the simulation platform,and the column inspection environment was built on the simulation platform.The forward and inverse kinematics simulation of the robot was carried out,which was compared with the real robot system to verify the correctness of the calculation results.At the same time,the spatial path planning simulation and collision detection simulation based on dynamic viewpoint of the train inspection robot were carried out.By comparing with the traditional AABB bounding box collision detection method,the high efficiency and rapidness of the algorithm were verified.