Multiple-model adaptive control architecture for a quadrotor with constant unknown mass and inertia

Abstract This paper proposes an architecture based on a multiple-model solution for height and yaw control of a quadrotor transporting an unknown constant load, added before the flight. Estimates of the inertial parameters (mass and z-axis inertia) and state variables (vertical position, velocity, yaw angle and rate) are obtained using data from the onboard sensors. A Multiple-Model Adaptive Controller (MMAC) architecture is proposed. The control of each partial model is based on a steady state Linear Quadratic Regulator (LQR), using integrative action for the height control. The resulting system is validated with load variations of up to 10% of the vehicle mass, both in simulation and experimentally, resorting to an off-the-shelf commercially available quadrotor.