Abstract: To deal with hovering control of four-rotor UAV from a practical point of view, a scheme of multi-sensor data fusion and a corresponding law of cooperative control are proposed. Based on the overall hardware frame of the aircraft, and within-depth analysis of each sensor in the measurement system, a Kalman filtering algorithm is developed for attitude measurement and a complementary filtering algorithm developed for height measurement. A dual sensor fusion algorithm is proposed to measure horizontal position. A four-rotor UAV model is established using small perturbation linearization. Based on the model, a corresponding cooperative control algorithm is designed. Simulation is performed to test the practicability. All algorithms are applied to a physical platform to verify effectiveness. The results show that the multi-sensor fusion algorithm combined with a coordination controller is reliable and effective. It can effectively improve accuracy of fixed-point hovering of four-rotor UAVs.

Key words: four-rotor UAV, data fusion, Kalman filter, complementary filter, cooperative control