Improve RRT Algorithm for Path Planning in Complex Environments

This paper proposes an improved RRT algorithm, which overcomes the problems of non-optimal path and low planning success rate in complex environments. First, the algorithm introduces a memory strategy that avoids the problem of oversampling or falling into local minimum in the non-convex environment during the process of expanding random tree. Then, the heuristic function is used to guide the growth of random tree and neighborhood expansion strategy is used to avoid the "rewiring" process of RRT* algorithm, which improves the real-time planning performance. Finally, simulation experiments are performed in different types of maps to illustrate the effectiveness of the proposed algorithm