System Identification and its Application to Estimating Soil Properties

Abstract

This paper examines the meaning of principal methods of system identification (SI). System identification is seen as a unique procedure to estimate in situ soil properties, especially soil subject to large strain. When applicable, parametric modeling of system processes is found to be superior to traditional Fourier methods. Several different methods have been presented to directly assess nonstationary data. These methods cover a wide range, from transforming the data into a stationary signal to full-fledged nonlinear, nonstationary analysis. The method used will depend on the nature of the data available and the nature of the requisite information. Segmentation of the data into stationary pieces gives sound results and is widely used. However, if the process of interest has time-varying parameters, a recursive technique should be used.