Modelling and constrained predictive control of a 3DOF helicopter

Predictive control methods have become widespread in industrial practice, but their applications have generally focused strictly stable systems with slow dynamics. In this paper, a state-space predictive control law is implemented for a three-degree-of-freedom helicopter with two input variables. The system has nonlinear, 6th order dynamics with very small damping. By linearizing a nominal physical model around an equilibrium point, a marginally stable prediction model was obtained. The real-time implementation involves the solution of a Quadratic Programming problem at each 50ms sampling period. Experimental results are employed to investigate the effect of changes on the prediction and control horizons, as well as to illustrate the ability of the controller to enforce output constraints. Keywords— predictive control, control with constraints, quadratic programming, real-time control, multivariable systems.