A digital twin framework for air-breathing precooled engine experimental system

Air-breathing precooled engines are pivotal for hypersonic flight but face challenges in experimental testing due to high costs, risks, and complex parameter coupling. This study pioneers a digital twin framework for such engine test systems. The framework integrates physical experimental facilities, a digital experimental facility, an information transmission system, and a data-knowledge system. The digital twin achieves reliable physical-virtual interaction and real-time state mapping through multi-disciplinary dynamic simulation. An integrated LSTM-based module enables accurate, real-time prediction of critical parameters and performance degradation. Validations show that the dynamic simulation maintains relative errors below 8 % for key thermodynamic parameters and thrust at typical speeds, while the precooler temperature prediction error is under 1.5 %. Thrust prediction remains accurate across diverse operations, including steady state, transients, and cold start. The information transmission system exhibits a low latency of ∼50 ms. Overall, this digital twin establishes a novel system for the efficient, safe operation and intelligent optimization of complex experimental systems, significantly reducing testing time, cost, and risk.