A multiobjective optimization issue: genetic control planning or trajectories.

Abstract

Part of a research project on cooperative marine robotics is the scenario of a submarine rendez-vous. This paper considers this case, where a high-manoeuvrability AUV (autonomous underwater vehicle) should meet a submarine platform for energy, samples and data service. Since the AUV is equipped with a set of thrusters, the problem of an adequate command of the thrusters appears. Given initial and final points for the AUV underwater trajectory, the question is to determine the set of forces and times to be exerted by the thrusters to get an adequate trajectory. Several constraints and simultaneous objectives to be optimized must be considered. Given the complexity of the multi-objective optimisation problem, it seems opportune to use Genetic Algorithms. The paper describes the problem to be solved, then explains how the GA were applied, and presents results for a set of cases considered, including obstacle avoidance.