Sparse Identification of Nonlinear Dynamics (SINDy) for Digital Twinning and Performance Modeling in Hybrid Engines

Abstract The advancement of digital twin technology is crucial for optimizing the performance, reliability, and efficiency of hybrid turbo-shaft engines in aerospace applications. This study explores the Sparse Identification of Nonlinear Dynamics (SINDy) framework for developing computationally efficient models that capture key engine dynamics using synthetic data and identify governing equations for critical performance parameters such as compressor exit pressure, shaft speed, turbine inlet temperature, fuel flow rate, turbine torque, and power lever angle. The proposed approach achieves high predictive accuracy, demonstrating its effectiveness in modeling complex turbomachinery behavior. Our results highlight the potential of SINDy for performance monitoring, anomaly detection, and predictive maintenance in aerospace propulsion systems.