Wide-range model predictive control for aero-engine transient state

Abstract

To perform transient state control of an aero-engine, a structure that combines linear controller and min–max selector is widely adopted, which is inherently conservative and therefore limits the fulfillment of the engine potential. Model predictive control is a new control method that has vast application prospects in the field of aero-engine control. Therefore, this paper proposes a wide-range model predictive controller that can control the engine over a wide range within the flight envelope. This paper first introduces the engine parameters and the model prediction algorithm used by the controller. Then a wide-range model prediction controller with a three-layer nested structure is presented. These three layers of the structure are univariate controller, nominal point controller, and wide-range controller from inside to outside. Finally, by analyzing and verifying the effectiveness of the univariate controller for small-range variations and the wide-range model predictive controller for large-range parameter variations, it is demonstrated that the controller can schedule the controller’s output based on inlet altitude, Mach number, and low-pressure shaft corrected speed, and ensure that the limits are not exceeded. It is concluded that the designed wide-range model predictive controller has good dynamic effect and safety.