Efficient Path Planning and Collision Avoidance for Autonomous Robot Navigation

The motion planning problem is a fundamental challenge in the field of robotics. This study presents a novel sampling-based pathfinding algorithm that enables robot to navigate complicated terrain by providing obstacle avoidance and local trajectory planning. Sampling-based pathfinding algorithms have been extensively utilized in motion planning challenges. The method has several limitations, including a sluggish convergence rate, significant variation in search time, a vast and dense sample space, and non-smooth search routes. This study introduces RE-RRT* (Robust and Efficient RRT*), an enhanced version of the sampling-based pathfinding algorithm designed to efficiently discover a rapid path that is nearly optimal. RE-RRT* iteratively improves the path by utilizing the Choose Parent and Rewire processes. As the random tree expands, the sample space becomes restricted, resulting in a reduction in unnecessary searches. Simulation demonstrate that the RE-RRT<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">*</sup> algorithm is capable of achieving smoother and more efficient convergence to a shorter path with fewer iterations, even in the presence of various obstacles. The proposed approach can enhance search efficacy, expedite convergence, and reduce processing time. Due to these factors, our proposed RE-RRT<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">*</sup> algorithm has superior performance compared to previous sample-based algorithms in terms of stability, speed, computing time, and sampling space.