324 Acquisition of Stable Flying Behavior under the Virtual Air Environment

Abstract

This study aims at an acquisition of the stable flying under the virtual air environment that models the drag and the buoyancy.The designed model is inspired by existing creatures with several wings, and controlled by the artificial neural network.Simulations by these models are carried out to evolve the proper behaviors by genetic algorithm.Simulation results verify that these models can fly stably by using wingstroke and gliding under the virtual air environment.