Research Of Six-dof Manipulator's Path Planning

In the robot's intelligent control, obstacle avoidance path planning is a key and difficult problem. Its mission is able to be stated as follows: in obstacles 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 start point to destination point, plans an optimal (or sub-optimal) collision free (obstacle avoidance) path. According to the structural features of the six-DOF repairing manipulator and the narrow nature of the work environment (steam generator), an obstacle avoidance path planning method is given, which is based on the RRT algorithm.Before the research of the manipulator obstacle avoidance path planning, a kinematic model and solving methods for the manipulator's working model are given. Among them, inverse kinematic problem-solving is more complicated. According to the low redundancy, limited working environment of the manipulator, a simple and easy to apply geometry diagram method is given, and through simulation experiments to verify it.An in-depth discussion for the collision detection is given in the paper, when the manipulator is working. And in accordance with the characteristics of the control system, a grid bounding method for the collision detection is introduced, which is based on the AABB box and equivalent cylinder. This method is simple and effective and can solve the collision detection of the path planning well.A discussion for the RRT algorithm is given, about the sampling strategy, expansion strategy, efficient finding nearest points and the number of the search tree etc. A balanced bidirectional search tree method for the path planning of the manipulator control system is given. VC++ development platform is applied. The simulation program of manipulator obstacle avoidance path planning is work out, and a detailed analysis method of its software design and implementation. At the same time, the simulation experiments have been done in the manipulator control system. The feasibility and availability of the algorithm are verified by the comprehensive analysis of the results.