An Evolutionary Approach to Tuning a Multi-Agent System for Autonomous Adaptive Control of a Flapping-Wing Micro Air Vehicle

Abstract

Biomimetic flapping wing vehicles have attracted recent interest because of their numerous potential military and civilian applications. In this paper, we describe an evolutionary approach to tuning a Multi-Agent System for autonomous adaptive control of a Flapping-Wing Micro Air Vehicle. The wings of the vehicle are controlled by a split cycle oscillator, which combined with non-linearities and differences between each vehicle, brings significant challenge for selecting the proper parameters for the control system. Adopting a Neo-Darwinistic evolutionary approach, where solutions are evolved in a similar manner as in nature, allows us to precisely learn control parameters for each vehicle. After describing the evolution algorithm and evolving the control parameters, we utilize these values for autonomous waypoint following by the micro air vehicle.