Search multi-unmanned ship path planning algorithm
Longxiang Wang, Kai Sun, Fan Tang, Yuanpeng Song, Chuyu Wang, Xu Feng, Fengbo Zhou, Ziteng Guo, Zeyang Wang · 2024
This paper proposes a dynamic CBS (DCBS) algorithm to address the limitations of existing conflict-based search (CBS) algorithms and their variants. CBS algorithms have been found to have prolonged search times and planned paths that do not align with kinematic constraints. The DCBS algorithm aims to address these issues by incorporating a dynamic approach to CBS. A suboptimal factor is introduced to enhance the search efficiency of CBS. Additionally, the A* path search within the CBS algorithm is enhanced to a hybrid A* search, thereby ensuring that the planned paths are more compatible with the kinematic constraints. Furthermore, a path distance weighting factor and a time weighting factor are incorporated into the underlying path search of the CBS algorithm to facilitate the planning of a fast and short path. Simulation experiments were conducted on the ROS platform for comparison purposes. The results demonstrated that the paths planned by the proposed DCBS algorithm were more compliant with the kinematic constraints of the unmanned ships than those planned by the cutting-edge ECBS algorithm. Additionally, the speed of path planning was improved by 47.06% on average, and the total path cost was reduced by 7.64% on average. These improvements in efficiency can be attributed to the effectiveness of the proposed DCBS algorithm in multi-unmanned ships operations.