Abstract
Flutter is destructive self-excited oscillation, which occurs when wind power is infused into aeronautical/industrial structures such as aircrafts and suspension bridges. Flapping power generator is a system which extracts wind energy from it. Recently, Isogai et al. proposed a new system. In his system, wing is supported elastically in heaving oscillation while the pitching oscillation of the whole wing is mechanically driven by an electric motor with a prescribed frequency and pitch amplitude. This system is governed by six non-dimensional parameters. In his study, these design parameters were optimized to maximize power efficiency by using "Complex Method". This paper considers multi-objective optimization of Isogai's "Flapping Wing Power Generator". In this study, objective functions are efficiency and power. To obtain a wide variety of Pareto solutions efficiently, one of Evolutionary Algorithms, Adaptive Neighboring Search (ANS), has been extended for the present optimization. In result we obtained unique tradeoff from the non-dominated solutions and solutions as good as by "Complex Method".
