Modeling, Identification, and Control of Non-minimum Phase Dynamics of Bi-copter UAVs

Bi-copter UAVs have emerged as more agile and energetically efficient flying platforms than quadrotor UAVs. Yet, they are more challenging to control due to the use of servo motors leading to non-minimum phase roll dynamics. In this paper, we systematically investigate the modeling, identification, and control of such non-minimum phase dynamics in an integrated way. We start with deriving the first principle model, from which the cause of the non-minimum phase is clearly seen. Parameters of the derived model are identified via experiments and subsequently used to design and analyze a high-gain robust flight controller via H-infinity loop shaping techniques. The designed controller is implemented on actual systems and compared against a well-tuned PID controller. Extensive flight tests are conducted to verify the controller's effectiveness and the effect of the system's non-minimum phase nature.