Effective emission control of aero-engines via nonlinear dual-estimators for uncertain states and parameters

This paper proposes an effective emission control strategy for an aero-engine using nonlinear dual-estimators, which aims to address challenges arising from state and parameter uncertainties. Transient air-fuel ratio (AFR) regulation issues, primarily caused by the wall-wetting and manifold-filling phenomena, are mitigated by a novel approach leveraging dual nonlinear estimators: (i) an extended Kalman filter (EKF) and (ii) an unknown dynamics estimator (UDE). The EKF estimates the fuel mass flow rate and unknown internal parameters, while the UDE compensates for nonlinear air-filling dynamics by estimating a lumped term involving the second derivative of air mass flow. The control framework is built upon a modified mean value engine model (MVEM), tailored to capture the unique dynamics of the rotary aero-engine. Building upon the foundational concepts developed in earlier research, this work moves toward practical application by demonstrating the proposed strategy through real-world experiments conducted on an AIE 225CS rotary aero-engine under a standard driving cycle, marking a significant step forward from theory to application. Furthermore, it introduces a complete stability analysis that addresses the coupling of dual-estimation errors, an aspect not explored in prior work. Comparative experimental results against an industry-standard gain-scheduling PID controller demonstrate the proposed method's performance in achieving improved AFR regulation and significant emissions reduction.