Adaptive nonlinear predictive control for turboshaft engine

When the collective pitch of the helicopter main rotor changes,in order to keep the speed of power turbine constant,and improve the dynamic performance,a novel adaptive nonlinear predictive control (ANMPC) algorithm was studied for turboshaft engine control system.Based on the steady states data and the dynamic properties of turboshaft engine,using recursive least square method to realize parameter identification,a numerical-ARX(auto regressive with external input) parallel prediction model which possessed online adaptive ability,was constructed.Then,by multistep output prediction and feedback correction,employing sequential quadratic programming,a ANMPC controller for turboshaft engine was obtained on the basis of receding horizon optimization.The simulation results show that,with the changing of collective pitch,ANMPC controller can make the power turbine speed converge faster,and make overshoot/droop smaller,compared with the conventional series PID(proportional-integral-derivative) controller.