Path planning of Autonomous Underwater Vehicles for optimal environmental sampling

Abstract

Dynamic variability in vast ocean environments and limitation in sampling resources have made optimal environmental sampling a hot topic of discussion recently. In this paper we consider the problem of path planning for Autonomous Underwater Vehicles (AUVs) carrying conductivity, temperature and depth (CTD) sensors, aiming to minimizing sound velocity profile prediction uncertainty after assimilating in-situ measurements. The problem is modeled as a non-linear deterministic optimization problem based on maximum a posterior probability (MAP) criterion. An approximated way to calculate the uncertainty reduction for each path group is utilized to save the computation time. Numeric simulation results have confirmed the effectiveness of the approach.