A multilayer path planner for a USV under complex marine environments

In this paper, a multilayer path planner (MPP) with global path-planning (GPP), collision avoidance (CA) and routine correction (RC) for an unmanned surface vehicle (USV) under complex marine environments including both coastal and surface constraints is presented. The main contributions of this paper are as follow: 1) An MPP framework consisting of multiple layers, i.e., backbone, CA and RC, is established, and achieves self-tuning path planning which adapts time-varying environments. 2) To minimize yaw-cost for the USV within local path, a novel CA algorithm is developed by the B-Spline method. 3) For capturing environmental influences arisen from reefs around the coastline, a stochastic dynamic coastal environments (SDCE) model is built by virtue of Poisson distribution. In combination with the fast marching method (FMM) and the SCDE model, the RC algorithm is proposed to handle environmental uncertainties. Simulation results show that the proposed MPP achieves remarkable path-planning performance in terms of both collision avoidance and adaptability to complex environments.