Quasi-Random, Maneoeuvre-Based Motion Planning Algorithm for Autonomous Underwater Vehicles

Authors:	Tan Chiew Seon, The University of Plymouth, United Kingdom Sutton Robert, The University of Plymouth, United Kingdom Chudley John, The University of Plymouth, United Kingdom
Topic:	7.2 Marine Systems
Session:	Marine Systems I
Keywords:	Hybrid system, Manoeuvre Automaton, Dynamics quantisation, Optimal motion planning, Motion primitives

Abstract

This paper presents an approach using a hybrid modelling techniqueknown as Manoeuvre Automaton (MA) to capture the key dynamics of a nonlinearautonomous underwater vehicle (AUV) in such a way that high-level tasks suchas optimal motion planning can be computationally simplified while still allowingit to perform complicated manoeuvres when the situation arises. With respectto motion planning in an obstacle filled environment, an incremental stochastictechnique derived from the Rapid-exploring Random Tree (RRT) algorithm isapplied. This paper proposes a multiple nested node version of RRT and alsoaddresses the case of a time varying final state. Simulation results as presented,using a 3 degree-of-freedom (DOF) AUV model in order to prove the viability ofthe concept. Copyright c° 2005 IFAC