RTRI REPORT May 2001

Analysis of Vehicle Vibration Using Linear Prediction Model

Tadao TAKIGAMI, Takahiro TOMIOKA


  In this paper, two linear prediction models, AR and ARMA models, are employed to analyze vehicle vibration characteristics. First, the authors calculate the frequency response function (FRF) and power spectral density (PSD) based on the AR model without using the conventional FFT algorithm. The ARMA model is then utilized to identify the vibration shape, or modal characteristics, such as the natural frequency, damping ratio and mode shape function. Analysis results show that the AR and ARMA models are suitable for vibration analysis even if the time to collect the identification data is not long enough for FFT calculation.


* Full papers of RTRI Report are written in Japanese.

Copyright (c) 2001 Railway Technical Research Institute