Feedback linearization control of multivariable and nonlinear systems (UAV)

Abstract

ors have been widely used for many applications; furthermore, various techniques for their modelling and control have been proposed. Among the challenges encountered in the design of controllers for a quadrotor is the fact that it is a highly coupled and nonlinear multivariable system. It is also nown as being an under-actuated system because it uses four actuators to control six degrees of freedom. In this work, the nonlinear dynamic model of the quadrotor is formulated using the Newton-Euler method. Then different linear and nonlinear control techniques for quadrotor trajectory tracking are investigated. First, SMC and PD controllers for linear and nonlinear trajectory tracking of the quadrotor are implemented, and genetic algorithm is used to optimize the controller parameters according to different objective functions. Both techniques are evaluated and compared in terms of trajectory tracking capabilities, dynamic performance, and the effect of possible disturbances. Then, backstepping and gain scheduling ontrol techniques are designed in order to control the altitude and attitude of the quadrotor, in the absence of disturbances and also in a windy environment. Finally, a new feedback nearization approach based on coordinate transformation and state feedback, is proposed. In this approach, the state space description of nonlinear quadrotor system is transformed into a linear quasi block controller decoupled form, then eigenstructure assignment using state feedback is applied. The proposed approach is used to control a quadrotor, in order to assess its performance in terms of trajectory tracking capabilities, time response performance, robustness and robust stability. The simulation is carried out using MATLAB/Simulink software.