Abstract—This paper presents a novel anti-disturbance tracking control problem for Unmanned Aerial Vehicles(UAVs) systems. Different from the general external disturbances, this paper uses the neural network to describe the nonlinear disturbances. On this basis, a disturbance observer is designed to estimate the nonlinear disturbances. By integrating the estimation of disturbance with P-I control algorithm, a composite controller based on convex optimization theory is designed to ensure the UAVs system stability and convergence of the tracking error to zero. Finally, Matlab / Simulink is used to simulate a Unmanned Aerial Vehicles model, which helps to verify the feasibility and effectiveness of this algorithm.