Adaptive Combustion Regulation in High-Fidelity Computational Model of Solid Fuel Ramjet

Abstract

Controlling the combustion process under hypersonic conditions remains a significant challenge. This paper uses a data-driven, learning-based control technique to regulate the combustion process within a solid fuel ramjet, aiming to regulate the generated thrust under uncertain operating conditions. A high-fidelity computational model combining compressible flow theory with equilibrium chemistry is developed to simulate combustion dynamics. This model evaluates the stability of the combustion dynamics and defines the engine’s operational envelope. An online learning controller based on retrospective cost optimization is integrated with the computational model to regulate the thrust. Numerical simulations indicate that the learning control system can regulate the thrust generated by an SFRJ without requiring any modeling information.