Multi-objective shape optimization of autonomous underwater glider based on fast elitist non-dominated sorting genetic algorithm

Abstract

Autonomous underwater glider (AUG) equips with limited battery capacity, and needs to optimize the shape of AUG to reduce power consumption and improve voyage. This paper presents a new method of the multi-objective optimization of AUG shape based on the fast elitist non-dominated sorting genetic algorithm (NSGA – II). The slender ellipsoid line is chosen as the reference model and the volume of the model is constrained to keep 100 L. The hull drag and the hull surface pressure are two key technical performance indicators. Variables are used for sensitivity analysis based on One-At-a-time (OAT) method. Comparisons between towing tank experiments and numerical simulation method is conducted to prove that this method is used for hydrodynamic analysis. The original shape, the NSGA-II optimization shape, the Spray shape and the multi-island genetic algorithm (MIGA) optimization shape are analyzed to verify the validity of the optimization method in this paper by comparing hydrodynamic performance and power conversion efficiency. The simulation results indicate that the NSGA-II shape obtains a better hydrodynamic performance than the others. At the same wing configuration and gliding depth, the voyage of the NSGA-II shape is more than the original shape 12%, which has great significance for reducing power consumption.