Analysis And Design On 6-dof Manipulator Path Planning

Robot's path planning plays a very important role in the control of robotics. It has universal applications in many fields, such as robotics, design of Very Large Scale Integrated circuits (VLSI), GIS, etc. The main contents relate to many subjects which includes environment representing, planning method, path searching and artificial intelligence. Researchers are always caring for much of the study of high-powered algorithm for avoidance obstacle path planning. Some achievements have been obtained through exploration and research in many aspects, but there are plenty of problems needed to be lucubrated. Based on demands and restrictions of PUMA 560 manipulator, this paper presents two models for the optimization of collision avoidance path planning.Firstly, the thesis introduces the theoretic foundation and the kinematics of six degree of freedom (DOF) manipulator. Then the inverse kinematics solution of the manipulator is brought forward on the basis of PUMA 560. And in the third part, several typical path planning algorithm are analyzed, which contains the artificial potential field, the Genetic Algorithm/GA and Fuzzy Logic/FL. Avoidance obstacle path planning problem is able to be stated as follows:in obstacle environment, based on a certain evaluation criterion, such as the shortest length of path, the shortest time of moving, the minimal consuming of energy and so on, from the start point to the destination point, plans a optimal(or sub-optimal)collision free path. Two models are proposed for the six-DOF collision avoidance.In model one, using the method of regular enveloping modeling of obstacles, and projecting obstacle and manipulator to two planes, the three-dimensional problem is transformed to a two-dimensional one and then the path is searched by A* algorithm. It greatly reduces the calculation and meets the demands of path planning.The first part of model two constructs the workable configuration space of the manipulator by Monte Carlo algorithm. The space is called free zone. Then, the information which denotes the minimum energy consumption and the maximum distance between manipulators and obstacles is appended on the edges of the roadmap which is set by the inverse Breadth-first Search; thereby the road map is transformed to network. At last, the optimization model is represented to ascertain a safe motion based on the network.In the last part, the effectiveness and safety of the proposed models are evaluated through computer simulations of motion planning for PUMA 560 manipulator, which shows that the algorithms in the two models satisfy the demands in different working conditions.