DETERMINATION OF KANAI AND TAJIMI PARAMETER USING ADAPTIVE FILTER

Abstract

Kanai – Tajimi parameters are used to characterize the strong ground motion. These parameters for different earthquake data are estimated from power spectral density vs frequency graph using spectral moments. In this thesis first earthquake signal is predicted from Adaptive filter then PSD vs frequency graph is obtained using low pass filter. In order to monitor, the change with earthquake shaking, the Kanai –Tajimi parameters (i.e natural frequency and damping ratio of soil layer) are also identified based on windowed data of earthquake using a moving time window. This thesis provides a method for estimating dynamic variations of K-T frequency and damping ratio parameter of soil related very closely to the nonliner earthquake responses of ground. It is showed that nonlinear responses with great reduction in soil natural frequency and increase of soil damping ratio occurred during the strong ground motion. Such a reduction of K-T frequency parameter and increase in K-T damping ratio parameter was found to have time – varying characteristics, so the further topic of this study focused for analyzing non-stationary variation of K-T parameter.