Research On Flight Control Algorithm Of Quadrotor Unmanned Aerial Vehicle

Because of the advantages of good flexibility, safe and cheap, the quadrotor unmanned aerial vehicle(UAV) has been widely applied in the aerial photo, disaster rescue, counter-terrorist surveillance, border patrol and other fields. Facing with the complex working environment, an efficient control system and a precise attitude estimation filter are the key to complete a task. However, the quadrotor's characteristics of nonlinear, strongly coupling and under actuated, increase the difficulty on designing the control system and state estimation filter. The paper analyzes the latest research and the key technology on quadrotor, completes the research work of modeling, control algorithm and attitude estimation algorithm.Firstly, the dynamic analysis of the quadrotor is done. Base on the Newton-Euler equation, the comprehensive dynamic model of the quadrotor is established. To avoid appearing singularity the model uses quaternion for attitude calculation. The model takes the translation and turn air resistance into account, and considers the motor rotors, propellers and quadrotor body as a multi-rigid-body system. The inertial parameters are extracted from Solidworks software.Then, combining with quadrotor's simulation model, the attitude and position's PID control strategy on two attitude feedback form(the rotation matrix and quaternion) is proposed. And the corresponding attitude control system is established. By proposing the corresponding transformation strategy from position error to attitude error, the cascade position control system is established. Simulation results show that the controller with quaternion feedback has better performance, and has advantages of simplicity, non-singularity and less calculation.Again, based on the quaternion feedback quadrotor model, the Active Disturbance Rejection Control Technique(ADRC) algorithm is proposed. With deducing the quaternion integrator series equation and combining the meaning of the attitude error quaternion, the ADRC control system of quaternion vector form is established. Simulation results show that the new control system has faster response speed, stronger robustness.Finally, based on UKF filter principle of nonlinear system, a motion filter of UAV is designed to preprocess sensors' data. Then a multi-sensor complementary fusion filter on accelerometer, magnetometer and gyroscope is proposed. The filter successfully estimates the attitude of UAV in the simulation environment.