Research On Obstacle Avoidance Motion Planning Method For Spatial Multi DOF Serial Manipulator

Robot technology has gradually become one of the main features of the intelligent manufacturing industry in the double background of industrial 4.0 and Chinese manufacturing 2025.At present,different kinds of robots,especially manipulators,has been involved in a lot of fields like industrial manufacturing,aerospace,entertainment services,medical,military and other fields.Moreover,spatial multi DOF serial manipulator is widely used in these fields and the research on its kinematics is one of research highlights.Above all,manipulator’s path planning technology is one of key technologies in robot’s motion control and it effects the properties of manipulator’s kinematics directly.However,a complicate environment with multi-obstacles severely limits the application and development of robot arm path planning technology.For this reason,study on the obstacle-avoidance path planning technologies of the spatial multi DOF serial manipulator has its practical significance and research value.This thesis mainly includes the following four parts:Firstly,mathematical model of a general 6-DOF serial manipulator is constructed by D-H theory and its function of forward kinematics and inverse kinematics are derived.Secondly,three common enveloping models of spatial geometry are proposed based on the problem of avoiding collision in the environmental with obstacles during the path planning of manipulator.They are the cylindrical envelop model,the rectangular envelope model and the sphere envelope model.Then,by simplifying the manipulator’s model,the problem of the joint interference of the manipulator and the collision detection between the manipulator and the obstacle is transformed into the problem of the identification of the relative positions of spatial point-line,spatial line-line,and spatial line-plane.At last,the problem would be solved by spatial analytic geometry knowledge.Thirdly,two methods are proposed to solve manipulator’s obstacle avoidance problem.One is based on the A * algorithm and the manipulator’s inverse kinematics in Cartesian space,which only applies to manipulators that satisfy Pieper rule;the other is based on the improved RRT algorithm and the manipulator’s forward kinematics in joint space,which establishes the cost function with the requirement of the end-effectors’ motion smoothness and trajectory optimization and is applicable to any articulated DOF serial manipulator.Finally,in order to verify the theoretical results,a virtual prototype is constructed to carry out the simulation and a real manipulator is constructed for experiment.It shows that the two kinds of path planning methods can solve the manipulator’s obstacle avoidance optimal path planning problem from the starting point to the end point in multi-obstacles environment.