Road Roughness Evaluation by Curve-Fitting and Subspace-Identification Methods

In this paper, for a quarter-car model, root-mean square response variables are calculated for roads classified by the international standard organization, based on the straight-line spectrum models. Covariance expressions are derived for vehicle response variables assuming that the road spectrum is that of a Motor Industry Research Association (MIRA) spectrum and compared with the existing results in the literature. The international standard and the MIRA spectra are then compared with the models delivered by high resolution frequency-domain subspace identification algorithms. The IRI statistic of the quarter-car response is calculated for the road models built by the curve fitting and the identification methods from the measured road spectrum assuming that the road excitation is a zero-mean Gaussian process