RRT-A*-BT approach for optimal collision-free path planning for mobile robots

Abstract

This paper deals with the problem of optimal collision-free path planning for mobile robots evolving inside indoor cluttered environments. Addressing this challenge, a hybrid approach is proposed combining Rapidly-exploring Random Trees (RRT), A-Star (A*) and Back-Tracking (BT) algorithms (RRT-A*-BT). Thus, a vision system is used for a nearly-exact modeling of the environment through image processing. Moreover, each iteration of the basic RRT approach is guided by A* algorithm while trying to take the shortest path linking the robot current position to target . In case of a blockage, BT algorithm is used to get out the robot from this situation. Finally, Piecewise Cubic Hermite Interpolating Polynomial (PCHIP) is used to smooth the generated optimal path. RRT-A*-BT approach is validated through different scenarios; obtained results are compared with previous works on same environments with same conditions. The results prove that RRT-A*-BT is better and faster than other algorithms of the literature, such as Genetic Algorithms and Conventional RRT, in terms of (i) computation time,(ii) path length and (iii) transfer time..