In order to solve conservatism problem resulted from Min-Max selection structure currently implemented in aircraft engine control

a nonlinear model predictive controller was designed

by incorporating penalty functions into the nonlinear sequential quadratic programming (SQP) algorithm. Simulation results show that nonlinear predictive controller

whether outputs hit their limits or not

achieved faster response than Min-Max logic

particularly apparent when outputs reached their limits. In this case

the setting time is reduced from 1.16s to 0.82s

and almost no overshoot occurred.