Research On Modeling And Model Predictive Control For Quad-Rotor Helicopter

As the quad-rotor is a MIMO nonlinear coupling system, it is meaningful to have a deep study on it. The study about the modeling method and model predictive control (MPC) method of the quad-rotor is presented in this paper. The traditional modeling method of the quad-rotor is physical modeling. The model is got by analyzing the forces of the quad-rotor. As some assumptions are made and some nonlinear factors are ignored, the physical model is not very accuracy.As the physical model has such demerits, the identification modeling method is used to get the model of the quad-rotor. The ARX model, RBF-ARX model and the RBFRW-ARX model of the quad-rotor are presented in this paper. ARX model is a kind of linear models, which parameters can be got from the linear least squares method (LSM). RBF-ARX model and RBFRW-ARX model are a kind of state-dependent models which are varying according to the state of the system. The parameter estimation method (SNPOM) is applied to estimate the parameters of the RBF-ARX model and RBFRW-ARX model. The results show that these models of the quad-rotor have high precision.The model predictive controller is designed for the quad-rotor according to the ARX model, RBF-ARX model and RBFRW-ARX model. These models can be converted into the form of the state-space which can be used to forecast the outputs of the quad-rotor. The optimal inputs of the quad-rotor can be got by quadratic programming.The simulation control for the quad-rotor is realized with the model predictive control method based on ARX model, RBF-ARX model, RBFRW-ARX model. The results show the quad-rotor can reach to the expected position very quickly with the MPC method. The model predictive control method based on ARX model and RBF-ARX model can resist disturbance. The effect of control parameters in MPC method is also is studied.At last, the real-time experiment results of LQR control, model predictive control based on ARX model、RBF-ARX model、 RBFRW-ARX model are compared with each other. The results show that the model predictive control based on RBF-ARX model has the best control results in the step response, global response。The anti-disturbance experiments show that the model predictive control based on the RBFRW-ARX model has the best anti-disturbance ability.