Iterative Learning-Based Path and Speed Profile Optimization for an Unmanned Surface Vehicle

Most path-planning algorithms can generate a reasonable path by considering the kinematic characteristics of the vehicles and the obstacles in hydrographic survey activities. However, few studies consider the influence of vehicle dynamics, although excluding system dynamics may considerably damage the measurement accuracy especially when turning at high speed. In this study, an adaptive iterative learning algorithm is proposed to optimize the turning parameters, which accounts for the dynamic characteristics of unmanned surface vehicles (USVs). The resulting optimal turning radius and speed are used to generate the path and speed profiles. The simulation results show that the proposed path-smoothing and speed profile design algorithms can largely increase the path-following performance, which potentially can help to improve the measurement accuracy of various activities.