Real time path planning for a class of torpedo-type AUVs in unknown environment

Abstract

This paper presents a path planning scheme for a class of torpedo-type AUVs in unknown environments. Consider the case where an underactuated vehicle (similar to a carlike robot) is travelling in completely unknown environments. In this case, due to deficiency of necessary information, quite number of graph generation and corresponding graph searching methods cannot be directly applicable. In this paper, a VRBVG (virtual rubber band visibility graph) method is proposed to generate a visibility graph under the assumption that C <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">unknown</inf> (outside of the vehicle's sonar coverage) does not include any obstacle. Using this graph, a simple A* search algorithm with the travelling time taken as the heuristic cost function is applied to find a path. Since the environment is unknown, the proposed planning method can only provide kind of suboptimal path.