Improving Reduced-Order Rotating Detonation Engine Models with Data Assimilation and Machine Learning

The paper presents a method to enhance one‑dimensional rotating detonation engine (RDE) models by integrating data assimilation and machine learning. Continuous data assimilation (nudging) aligns the low‑order Koch‑Kutz solver with high‑fidelity temperature data, while a Jacobian‑regularized closure is trained on the recorded correction forces. The resulting corrected model, once the observation term is removed, autonomously reproduces the temperature spectrum and key statistical properties of the full high‑fidelity simulation.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.