Bezier Curve-based Motion Planning for Efficient Robot Exploration in Unknown Environments
Chaoxin Zheng, Jianbin Wu, Yuhang Bao, Changyun Wei · 2024
Planning an efficient trajectory remains challenging for an autonomous robot that explores an unknown environment with obstacles. In this paper, we propose a novel strategy using Bezier curves to further optimize the trajectory planned by the sampling-based approach. We develop an optimal Bezier path (OBP) algorithm to generate additional control points for avoiding the local obstacles while reducing the tortuous and redundant paths. Specifically, we prune the redundant waypoints suggested by the rapidly-exploring random tree (RRT) algorithm, and we use Bezier curves for smoothing the local spikes as well as optimizing the total trajectory length. The proposed method outperforms two baseline algorithms of NVBP and GBP in two environments using a simulated four-wheel unmanned ground vehicle (UGV) equipped with LiDAR.