A probability smoothing Bi-RRT path planning algorithm for indoor robot

Abstract

Bidirectional rapidly exploring random trees (Bi-RRTs) have been widely applied in path planning and have been demonstrated to yield the optimal path for indoor robots. However, the quality of the initial solution is not guaranteed, and the convergence speed to the optimal solution is slow. To overcome these limitations, this paper proposes a novel probability-smoothing Bi-RRT (PSBi-RRT) algorithm for path planning. Specifically, the kinematic model is established, and the posture of the indoor robot is estimated. Thereafter, a target biasing strategy is introduced to select sampling points in the sampling region. In addition, a node correction mechanism is applied to correct node coordinates to reduce the collision probability in cluttered environments. Finally, two trees are connected using the direct connection method. The θ-cut mechanism is designed to remove the redundancy points, and the remaining nodes are connected using a triangular inequality-based RRT to ensure that the final path is optimal. Furthermore, the proposed method benefits from the properties of Bi-RRT, which offers low-cost solutions with fewer iterations compared with Bi-RRT alone. Different simulations were performed to demonstrate the significance of PSBi-RRT in comparison with Bi-RRT, InformedBi-RRT*, and QuickBi-RRT*.