Path planning of unmanned surface vessel in an unknown environment based on improved D*Lite algorithm
Jiabin Yu,Meng Yang,Zhiyao Zhao,Xiaoyi Wang,Yuting Bai,Jiguang Wu,Jiping Xu
DOI: https://doi.org/10.1016/j.oceaneng.2022.112873
IF: 5
2022-12-15
Ocean Engineering
Abstract:To solve the problem of path planning for an unmanned surface vessel in an unknown environment, this paper proposes a path planning algorithm based on the improved D*Lite algorithm. First, to improve the computational efficiency of the traditional D*Lite algorithm, the path cost function is improved to reduce the expansion range of nodes, and the node expansion direction is limited to avoid double node computation. Second, the path planned by the traditional D*Lite algorithm on the grid map is suboptimal; so, the inverse distance weighted (IDW) interpolation method is applied to shorten the path length. To improve the smoothness of the planned path by the traditional D*Lite algorithm, the Dubins algorithm is introduced for local path smoothing, and a smooth collision-free path that conforms to the motion dynamics of the unmanned surface vessel is obtained. The proposed algorithm was verified by experiments in known and unknown environments. The performance of the proposed algorithm was compared with the traditional D*Lite algorithm in terms of planning time, path length, and path smoothness. The results revealed that the proposed algorithm has the shortest path length, a shorter planning time, and the best smooth path.
engineering, civil, ocean, marine,oceanography