Smooth Path Planning for Mobile Robot Based on Adaptive Rapidly-exploring Random Tree

Abstract

Path planning is a challenging issue in the field of mobile robotics. Path planner based on RRT is widely used due to their ability to find a feasible path rapidly and efficiently. Regardless of whether the environment is simple or complex, however, the traditional RRT algorithms always adapts fixed step length and bias probability to carry out path planning, which resulting in excessive redundant nodes and poor algorithm performance. An adaptive RRT algorithm is proposed which it can automatically adjust step length and bias probability according to the collision detection results. To obtain a shorter and smooth path, greedy approach is used to prune the path by deleting redundant nodes. Simulation results show that the proposed algorithm is highly competitive and can work efficiently.