The Study Of RRT Algorithm Based Motion Planning For Non-holonomic Mobile Robot

Path planning is one of the core contents of autonomous mobile robot programming, and there have been many achievements in this field. However, the state transition of path generated by traditional geometric path planning methods is too fast to trace for the mobile robots considering the differential constraint, the motion planning considering the non-holonomic differential constraints is becoming the hotspot and new challenge in the domain of path planning.The RRT algorithm is based on random sampling, meanwhile it is a single-query path planning algorithm. So the RRT algorithm is always to be applied to solve the motion planning problems in highly-constraints environment or the problems with high dimension.For the RRT algorithms using a bias towards the goal while choosing a random configuration will encounter the problem of local minima, this paper combines the mechanism of collision information detection into the RRT-Blossom algorithm, to help escape regions with local minima problem. The usage of mechanism of collision information takes the global constraints into consideration, meanwhile it avoids repeatedly exploring the directions which have been explored. The repeatedly regression testing and detection of obstacle avoidance to the edge nodes are also saved. The ultimate goal of the algorithm is to improve the efficiency of the planner, especially in the environment with highly-constraints.In the fourth part of this paper, an improved heuristic RRT algorithm is proposed to solve the motion planning problems for non-holonomic mobile robots. In this algorithm, the A*cost function is introduced into the RRT algorithm to optimize the performance of the path. Meanwhile, several different metric functions are applied as the heuristic information function. The simulation result of several improved heuristic algorithms verifies the excellent performance of improved RRT algorithm, and the difference of paths generated by heuristic RRT algorithm based on different heuristic information function is very large.