Unmanned Surface Vehicle Path Planning Based on Improved Informed-RRT* algorithm
Xuehang Lin, Enjiao Zhao, Xingshun Wang · 2023
Aiming at the problems encountered by the InformedRapidly-exploring Random Trees Star (IRRT*) algorithm in the research of unmanned surface vehicle path planning in electronic chart environment, an improved guided IRRT* algorithm is proposed. Electronic chart data are processed and raster into raster maps. The target-oriented sampling area is adopted to strengthen the global random sampling point grow in the target-oriented area, and the aimlessness global exploration problem in initial stage of path planning is solved. Introduce adaptive step size and growth angle constraints, calculate obstacles appearance frequency and the number of samples fall into local loop, automatically adjust the step size to accelerate exploration speed in sparse marine environments. Introduce ancestor pruning strategy to obtain an optimal path and improve the overall effectiveness of algorithm. Compared with original algorithm, the algorithm proposed in this paper has higher efficiency and shorter path length, especially in electronic chart environment. Simulation results show that the guided Informed-RRT* algorithm proposed in this article converts targets into target-areas, accelerate searching speed and improving the overall ability of the algorithm.