Intelligent ammonia precooling control for TBCC mode transition based on neural network improved equilibrium manifold expansion model

Abstract

Precise control of the aeroengine precooling temperature is important for safeguarding the mode transition (MT) in turbine-based combined cycle (TBCC) propulsion. Ammonia mass injection pre-compressor cooling (Ammonia MIPCC) technology increases the maximum operating Mach number of aeroengines, presenting challenges in the design of controllers for complex control objectives. This paper introduces the combined inlet characteristics and the precooling scheme in the MT and proposes a neural network-improved equilibrium manifold expansion (NNEME) model. An intelligent ammonia precooling controller for the ammonia MIPCC aeroengine (AMA) was established based on NNEME and extended state observer (ESO). Furthermore, the final error confidence intervals for the NNEME model show a significant reduction of 37.499 % for rotor speed and 37.919 % for compressor outlet temperature, compared to the conventional EME model. By implementing ESO-based online compensation for modeling uncertainties, we achieved remarkable improvements in the control results. Specifically, the maximum steady-state errors for rotor speed and compressor outlet temperature were reduced by 60.839 % and 51.529 %, respectively, when compared to control results without ESO integration. Therefore, the proposed NNEME + ESO-based controller enabled precise control of the AMA precooling temperature during the MT phase. This demonstrates the effectiveness of our approach in enhancing control accuracy and stability.