Prediction of engine combustion chamber outlet temperature field based on deep Learning: Application in aero-engine life extension control

Abstract

One of the current research hotspots is the prediction of combustion chamber flow field based on deep learning methods, but how to effectively utilize these prediction results is a key issue that needs to be addressed urgently. This paper further proposes a study on the predictive control method for engine life extension considering the dangerous point temperature of turbine guide vane based on the prediction of combustion chamber outlet temperature field. Firstly, a combustion chamber outlet temperature distribution prediction model is established through Computational Fluid Dynamics (CFD) numerical simulation and inception deconvolution network, and integrated into the variable cycle engine component level model to achieve real-time prediction of combustion chamber outlet temperature distribution. Furthermore, based on this, the dangerous point temperature of the high-pressure turbine (HT) guide vanes is introduced as a constraint into the prediction optimization, and a predictive control method for engine life extension considering the thermal mechanical fatigue life (TMFL) of the guide vanes is proposed. The simulation results show that compared to the traditional LEC control method, the temperature at the critical point of the blade is reduced by 12.28 K, and the TMFL of the blade is increased by 29.23 %. The research results of this article provide a good application example for predicting the temperature distribution at the outlet of the combustion chamber, expanding the application scenarios for predicting the flow field parameters inside the engine.