Research On Dynamic Motion Planning For Manipulator Among Movable Obstacles

Operating in the environment with movable obstacles is an operating common condition for the manipulator,it is the key to realize the safe operation of the manipulator in the dynamic obstacle environment that adjust the obstacle avoidance planning path in time according to the movement of the obstacle.when a 6R manipulator is working in a dynamic environment,there are problems such as large search space,poor real-time performance and low planning efficiency.Based on sampling planning and obstacle mapping,a dynamic obstacle avoidance planning method is proposed to solve these problems.Firstly,the workspace of manipulator is rasterized and the grid index is established.It is necessary to find neighboring nodes frequently in sampling planning,and in order to raise query efficiency a sample storage structure is designed.Then a mapping relationship between the task space and the joint configuration space is constructed.by querying the mapping relationship,online collision detection is avoided and the efficiency of obstacle avoidance planning is improved.Based on the fast marching tree algorithm,an inverse fast marching tree algorithm for offline planning and a dynamic fast marching tree algorithm for online planning are proposed.The performance and the effectiveness of the algorithm are verified in a two-dimensional map environment.Finally,the virtual simulation platform is built,the simulation experiment is completed,and the manipulator teaching system is developed.The simulation result is generated to the teaching system control code,which is applied to the obstacle avoidance operation under the real environment.It proved that the algorithm is useful for 6R manipulator's dynamic obstacle avoidance planning.In the first chapter,the problems existing in the operation of the manipulator in the environment containing moving obstacles are analyzed.the research status of the related techniques of the manipulator obstacle avoidance motion planning algorithm,planning space and obstacle expression are summarized.Then the research content of this paper is expounded and the organizational structure of this article is given.In the second chapter,the expression of the working environment of the manipulator combined with the task space and the joint configuration space is presented.According to the characteristics of the obstacle movable condition,the definition of the obstacle avoidance planning of the manipulator in the movable obstacle environment is given.Then,the HRT-120 manipulator is used as the research object,and the manipulator mathematical model is established.Monte Carlo method is used to simulate the task space of the manipulator,the task space is divided based on the grid method,and the grid index is designed.Finally,design data storage structure based on kd tree to accelerate the search efficiency.In the third chapter,an obstacle avoidance pre-analysis method for the construction of obstacle mapping relationship and real-time mapping of obstacles is proposed.Firstly,the simplification of the manipulator collision model and interference analysis are carried out for the characteristics of large computational complexity in the process of constructing the mapping relationship.Then sample and generate sample point sets and edge sets in the joint configuration space,construct the mapping relationship from the task space to the joint configuration space based on the traversal method.Finally,the dynamic obstacles are decomposed and mapped based on the surface sampling method and the hash table.In the fourth chapter,a dynamic obstacle avoidance planning algorithm based on fast marching tree is proposed.Aiming at the characteristics that the node growth information is closely related to the root node in the fast travel tree algorithm,an inverse fast travel tree algorithm is proposed.In the online planning process,the path information before the interference with the obstacle is fully utilized by separating and reconnecting the search tree,and the speed of replanning the path during the dynamic obstacle avoidance process is accelerated.In the fifth chapter,a manipulator teaching system is developed.Firstly,the virtual obstacle avoidance experiment platform is built based on SolidWorks and MATLAB.It is proved to be effective that the proposed algorithm applied to the 6R manipulator virtual experiment platform.The overall framework of the teaching system is introduced,and the teaching programming and teaching reproduction are emphasized.the obstacle avoidance planning module based on the teaching system is developed,the simulation result into the teaching code is converted,and the correctness of the result in the real environment is verified.The sixth chapter summarizes the research contents and innovation of this paper and looks forward to the follow-up research.