2 A Neuro-Evolutionary Approach to Micro Aerial Vehicle Control

This paper addresses Micro Aerial Vehicle (MAV) control by leveraging neuro-evolutionary techniques that accommodate a higher number of control surfaces. Applying classical control methods to MAVs is a difficult process due to the complexity of the control laws with fast and highly non-linear dynamics. These methods are mostly based on models that are difficult to obtain for dynamic and stochastic environments. Instead, we focus on segmenting the different control surfaces to allow more flexibility to the neuro-evolutionary based controller. Precise control is then achieved by neuro-evolutionary techniques that have been successfully applied in many domains with similar dynamics. The results show that MAV performances are improved both in terms of reduced deflection angles and reduced drag (up to 4%) over a simplified model in two sets of experiments with different objective functions.