Model Free Adaptive Predictive Control Method For Quadrotor Aircraft

The quadrotor aircraft has the advantages of simple control principle,diverse flight modes,flexible maneuverability,and high safety.Therefore,it has broad application prospects in civil and military fields.At the same time,the quadcopter is a complex system with strong coupling,multivariable,nonlinear and other characteristics.Among them,the attitude control part of the quadcopter is an over-actuated system with four inputs and three outputs,all of which bring difficulties to the attitude adjustment of the quadrotor.This thesis takes the quadrotor flight simulator as the research object,applies Model Free Adaptive Predictive Control(MFAPC)to the attitude adjustment problem of the quadrotor aircraft,and conducts a series of simulation experiments,the main content as follows:(1)Considering the predictive control introduces future input and output information and has strong robustness,combining the idea of predictive control with traditional model-free adaptive control,a model-free adaptive predictive control based on MIMO nonlinear system is proposed.This algorithm only uses the input and output information of the controlled system to design the controller,without any model information,and through rigorous theoretical analysis to prove the convergence of the algorithm.(2)In view of the characteristic that the quadrotor attitude control is a typical second-order system,the MFAPC algorithms is improved to increase the restriction on the tracking error change rate and the change value of the tracking error change rate.In the MATLAB / Simulink environment,the quadrotor adjustment attitude controller based on the MFAPC algorithm is modeled for subsequent simulation experiments.(3)The validity of the algorithm is verified by simulation.A series of numerical simulation experiments on the traditional two-input two-output system show that the MFAPC algorithm can make better control decisions within a period of time due to the future input and output information is considered in the control law,so it has better tracking performance,and the algorithm can realize the decoupling of the system.The simulation experiment is carried out on the attitude adjustment of the four-rotor aircraft system,and the simulation analysis and comparison of the MFAPC algorithm and the MFAC algorithm under the conditions of no disturbance,external disturbance and internal disturbance are studied.Simulation results show that,compared with the MFAC algorithm,the MFAPC algorithm has stronger anti-interference ability and tracking ability.