A Harmonic-Regression ARMA Model for Rotating Machinery Dynamics Identification and Fault Detection Under Varying ConditionsNew

This paper addresses fault detection challenges in rotating machinery operating under varying conditions. The proposed Harmonic-Regression ARMA model combines harmonic regression components with autoregressive moving average structures to capture periodic and stochastic dynamics in vibration signals. The method mitigates limitations of purely rational spectrum assumptions in traditional ARMA-type approaches while offering better interpretability compared to deep learning alternatives. The research contributes to rotordynamics and mechanical fault diagnostics by providing a parametric signal processing framework for condition monitoring.