Research On Modeling And Control Of A Quadrotor Helicopter

Quadrotor helicopter,which is simple in mechanical structure,can achieve vertical takeoff and landing,stable hover and autonomous cruise.Due to its stable flight performance,easy operation and high security,it is widely applied in aerial photography,power lines inspection,farming and other fields.However,the system of the quadrotor is multivariable,non-linear and strongly coupled,making it difficult to be precisely controlled.In this paper,we design a micro quadrotor helicopter,concentrate on the modeling and control method of the quadrotor.The following work is carried out:The nonlinear dynamic model of quadrotor is built based on flight theory and spatial rigid motion equation,combined with specific object of the quadrotor.In order to obtain the more accurate attitude model,the system identification experiment according to the least-squares algorithm was designed.Then by taking the velocity along X-direction as the scheduling variable,a velocity model was transformed into a LPV form by Jacobian linearization method,which can reflects the dynamic characteristics of the system.An attitude H_? controller is designed by using the mixed sensitivity method,to minimize the tracking error and achieve robust performance.Based on the velocity LPV model,the polyhedron of the LPV model is obtained by convex decomposition technique,then the linear matrix inequality(LMI)technology for each vertex of a convex polyhedron is designed to meet H_? performance,finally,a gain-variable controller was synthesized via convex decomposition technology.The controllers we designed were verified by simulation experiments,outdoor flight experiment are performed on the attitude controller and achieved good performance.