Navigation And Control Of Quadrotor

Quadrotor is a new model of small UAVs that could take off and land vertically. It adjusts its position and attitude by varying four symmetrical propellers'speeds, with merits such as hovering ability, good maneuverability, and simplified structure.The paper first analyzed the flight principle and designed the structure of micro quadrotor by using carbon fiber. Then it continued to established the quadrotor model based on mechanical structure with 6 degrees of freedom, and simplified the model due to practical control issues. This paper designed PID attitude controller and Backstepping attitude controller. Through the simulation, we got the attitude response curves under several conditions. The simulation results prove the feasibility of two algorithms, and indicate that Backstepping method was better in the attitude control of quadrotor system which was nonlinear and highly coupling.Secondly, due to the requirements of quadrotor on heading and attitude information, this paper designed an attitude and heading reference system based on gyros, accelerators and magnetic sensors, and designed extend Kalman filter to estimate attitude and heading information. In the filter, the quaternion was chosen as state variables, gyros were used to update the stats variables, and accelerators and magnetic sensors were combined to establish measurement equation. In comparison with the traditional attitude estimation method, this algorithm can greatly reduce computational complexity and was more suitable to attitude estimation based on embedded microprocessor. Experiment results demonstrated that this type of algorithm was effective in attitude estimation.Last, the paper designed main controller and ground control system for quadrotor based on the systematic and functional requirements. Flight tests include hovering, attitude tracing and anti-interference hovering were conducted and proved to verify the effectiveness of the control algorithm.