Improved rapidly-exploring random tree algorithm application in unmanned surface vehicle local path planning

Abstract:

A path planning method based on improved RRT (Rapidly-Exploring Random Tree) algorithm was proposed in application to local path planning of USV (Unmanned Surface Vehicle). Aiming at the high speed and real-time control of USV, the inhibitory factor, limited angle and distance heuristic information were introduced into classical RRT algorithm, thus the selection of exploration points and growing points were modified, and the speed of the algorithm was improved. The excess navigation points in planning path were processed and smoothed considering the gyration performance to make the navigation distance shorter and meet the special demands of maneuver performance of USV. The experiment of local path planning was completed based on the environment model constructed by the process results of typical radar images in the sea and lake experiments. The experimental results showed that the suggested algorithm could rapidly complete the path search, the efficient of algorithm was improved and the distance of path was reduced, the planning path after optimization treatment could satisfy the planning system need. The suggested method can apply to USV local path planning.