This paper presents an online robust non-linear trajectory tracking control strategy for a quadrotor helicopter (UAV). In order to improve the system response, the automatic-adjustable integral action is integrated to the non-linear controller. Then, the controller coefficients are optimized adaptively in an online way by using an efficient technique called multiobjective particle swarm optimization. In particular, the control law is designed adaptively with respect to the uncertainty of the dynamic model parameters which include the position of the center of gravity of the quadrotor helicopter. A novel non-linear observer that exploits a particle swarm estimator (PSE) is used to estimate the system states. A stability analysis, as well as numerical simulations, are provided to verify the control design.
