Research On Quadrotor Flight Control Based On Extended Kalman Filter And Active Disturbance Rejection Technology

With the fast development of intelligent robots in recent years,multi-source information fusion and control technologies have remarkably advanced.Multi-rotor drones,termed as aerial robots,have been widely applied in photography,plant protection,inspection,mapping and other scenarios due to their good maneuverability.However,the measurement accuracy of system states and flight performance are still needed to beimproved from the aspects of anti-jamming capability and flight control accuracy.Therefore,the development of algorithms based on quaternion separated extended Kalman filter and active disturbance rejection control(ADRC)are carried out.The main content and innovations can be summarized as follows: First,according to the hardware structure and physical characteristics of the quadrotor,a sixdegree-of-freedom mathematical model of the quadrotor including roll,pitch,yaw,altitude,and horizontal positions is represented.Second,formeasurement sensor modules mounted on the body,corresponding signal processing is performed according to signal characteristics.Next,a quaternion-based separated extended Kalman filter(EKF)algorithm is designed to solve the attitude information,and an adaptive EKF algorithm is implemented to solve the threedimensional position.Last,considering the susceptibility of drones to environmental disturbances,the flight control algorithm was developed by using the theory of ADRC.In the attitude solution experiment,the estimated error of the pitch and roll angles is about 1.2 degrees,and the estimated error of the yaw angle is about 1.0 degrees.In the integrated navigation experiment,the error of the height estimation is less than 0.14 meters,and the error of the horizontal position estimation is less than 0.27 meters.In the flight control experiment,the error of attitude control is about 0.7 degrees,the error of altitude control is less than 0.1 meters,and the error of horizontal control is less than 0.15 meters.Experiments show that the designed quaternion-based separated EKF algorithm can reduce the effects of magnetic field interference and body vibration.The adaptive EKF algorithm can adjust the covariance matrix of measurement noise according to the environmental change,and effectively reduce the influence of signal drift of positioning system and the decline of positioning quality after being covered;At the same time,the flight control algorithm of the quadrotor aircraft based on the ADRC proves to reduce the error of attitude control and position control,with good robustness and anti-interference ability.