Modal Parameter Identification Using Classical Prony Method and AR Parameter Modeling

This document presents an implementation of modal parameter identification using the classical Prony method combined with AR parameter modeling. The classical Prony method serves as a fundamental signal processing technique for extracting modal parameters from vibration signals through exponential fitting of sampled data. AR parameter modeling employs autoregressive algorithms to characterize signal dynamics using linear prediction coefficients. By integrating these approaches, we achieve enhanced accuracy in identifying modal parameters while providing deeper insights into system dynamic behavior. The implementation typically involves: 1) Signal preprocessing and sampling rate configuration, 2) Prony algorithm execution for pole extraction using least-squares estimation, 3) AR model order selection via criteria like AIC or FPE, and 4) Cross-validation between methods for parameter consistency. This combined methodology proves effective for obtaining detailed modal characteristics including natural frequencies, damping ratios, and mode shapes, with practical applications in structural health monitoring and mechanical system analysis.