Minimum-time Motion-Planning of AUVs under Spatially Varying Ocean Currents

This paper presents a novel method for motion planning of AUVs in the presence of spatially varying ocean currents. The proposed approach uses the concept of Renormalization group theory (RGT) to handle spatially varying ocean currents by merging regions with homogeneous currents (i.e., with similar direction and magnitude). Such regions contain redundant samples and lead to an increased planning time. The removal of redundant samples from the homogeneous regions results in non-uniformly distributed samples for motion planning. Then, a novel variant of the A* search algorithm, which is a hybrid of both sample-based and grid-based methods, is developed to plan the minimum time path between the start and the goal position. The proposed algorithm is guaranteed to provide better quality solutions in terms of the time-cost compared to the standard A* algorithm. Moreover, removal of redundant samples leads to a significant reduction in computational complexity.