Direct thrust control for variable cycle engine based on fractional order PID-nonlinear model predictive control under off-nominal operation conditions

Abstract

With the development and complexity of aero-engines, aero-engine control has been transformed to provide proper thrust and limit protection function under the framework of multivariable control. The multi-operating modes and multi-control variables of the variable cycle engine (VCE) lead to the complex and changeable conversion between measurable variables and thrust, making the traditional indirect thrust control hard to ensure the accuracy and safety of thrust control in different operation conditions. In this paper, a direct thrust control method based on fractional order PID-nonlinear model predictive control (FOPID-NMPC) algorithm is proposed to achieve the tracking of thrust and the limitation of parameters of the VCE under nominal and off-nominal operation conditions. In view of the existing thrust estimator designs which are only suitable for nominal conditions, the individual differences and performance degradation are fully investigated. After setting the label which represents degradation mode and constructing the health index which represents degradation degree instead of estimating health parameters directly, a dual margin degradation pattern classifier is constructed based on random forest (RF), and a long short-term memory (LSTM) neural network-based thrust estimator is designed to jointly constitute the estimation module. Then, FOPID and NMPC are combined to ameliorate the control quality and ensure the accuracy and security of thrust control at different operating conditions. This method can escape from local optimum when solving multi-extremum optimization problems. The comparative simulation of the proposed FOPID-NMPC with PID-NMPC and NMPC is performed and the influence of controller parameters on thrust response is explored. Simulations show that the designed controller has better dynamic performance, and also has good steady-state performance and safety.