Towards Bio-Inspired Robotic Aircraft: Control Experiments on Flapping and Gliding Flight

THERE is a growing interest in the aerospace community in the development of robotic micro aerial vehicles (MAVs) to learn and mimic avian flight. MAVs fly in low-Reynolds-number regimes of 10 to 10, which corresponds to that of small birds or bats [1]. MAVs with wings equipped with multiple degrees of freedom such as flapping, wing twist, and sweep provide greater payload capability than insect-like MAVs and greater maneuverability than conventional fixed-wing aircraft. These MAVs can be used for intelligence gathering, surveillence, and reconnaissance missions in tightly constrained spaces such as forests and urban areas. Advances in actuators and control systems have led to development and analysis of articulated and flapping MAVs inspired by animals [2–5]. Birds and bats achieve remarkable stability and perform agile manuevers using their wings very effectively [2]. One of the goals of reverse-engineering animal flight is to learn more about the various aspects of avian flight such as stability, maneuverability, and control from the dynamics of MAV. From a controls standpoint, there are three major flight regimes or modes in animals: high-frequency flapping, low-frequency flapping, and gliding. Insects reside almost entirely in the high-frequency domain, utilizing small musculature to produce passive pitching dynamics over the course of several wingbeats [6]. For these situations, we can utilize averaging theorems to greatly aid control design around trim states such as hover condition [3, 7, 8]. Much bat and bird flight is low frequency enough that averaging theorems do not apply in practice and intrawingbeat effects are significant [9]. Additionally, steady-state equilibria (or approximations thereof) do not exist, which distinguishes this regime from the other two. Dietl and Leonard give examples of an analysis of this regime by