Research On Intelligent Motion Planning And Obstacle Avoidance Of Industrial Manipulators

The obstacle-free motion planning of multi-DOF(degree-of-freedom)serial industrial robots is a key and difficult point in the robotics research.In general,Industrial robots work in an environment with many obstacles.Motion planning helps their end effectors find an optimal obstacle-free path from the initial point to the target point.The planning result will directly affect the performance of kinematics and dynamics of industrial robots.Therefore,this paper proposes a novel motion planning method of a 7-DOF serial manipulator,which combines the improved RRT(Rapidly-exploring Random Tree)algorithm and inverse kinematics solution method.The main content of this paper is as follows.Firstly,the D-H method is used to represent and model a 7-DOF serial manipulator,and its forward kinematic equation is deduced.Secondly,in order to find the inverse kinematics solution corresponding to the path nodes generated during the motion planning in the workspace.A new approach CPABC based on the ABC(Artificial Bee Colony)algorithm is proposed.The simulation results show that the proposed approach is of higher quality than the ABC algorithm,its derived algorithm and other swarm intelligence algorithms.Thirdly,the AABB(Axis-aligned Bounding Box)level bounding box is used for the collision detection between a 7-DOF serial manipulator and obstacles in motion planning.The AABB's each face is perpendicular to one of the axis of the base coordinate frame.Its structure is simple and combined with the hierarchical bounding box technology can reduce the time consumed by collision detection.Finally,this paper proposes a motion planning method combining RRT* and RRT-Connect,which both are improved RRT(Rapidly-exploring Research Tree)algorithm.The proposed method constructs two path trees from the initial point and the target point respectively according to RRT-Connect,and the search process is similar to that of RRT*.It ensures that a better path can be obtained in a limited number of search times(ie,complete probability and asymptotically optimal).In order to verify the effectiveness of the method,simulation experiments are performed using MATLAB and ROS(Robot Operating System).The results show that the motion path planned by the proposed method is better than other RRT algorithms in finite time.