GENERATION OF ARTIFICIAL EARTHQUAKE MOTION RECORDS USING WAVELETS AND PRINCIPAL COMPONENT ANALYSIS

In the last century alone two to three million people died in earthquakes; More than 240 000 perish in Tangshan earthquake in China, 20 000 in the Izmit [Ikeya, 2004] 50 000 in India Bhuj earthquake, 100000 in Sumatra and 90000 in Pakistan earthquakes. Earthquake engineering has progressed to the stage where it is now computationally practical and desirable to perform a dynamic analysis of most civil engineering structures. Such an analysis requires the engineer to create an accurate analytical model of the structures as well as prescribe an earthquake input excitation. The design earthquake input excitation for the site under consideration is usually prescribed in the form of response spectra or in the form of an ensemble of artificial earthquake acceleration time histories. There is every need for generating artificial accelerograms since recorded accelerograms are very limited at site. This paper proposes five neural network based models for the generation of artificial earthquake and response spectra using wavelet transforms (WT) and principal component analysis (PCA) where the recorded accelerograms are limited at site of interest. The proposed model is compared with Lee and Han's model. The data for 25 earthquakes are taken for training and 4 for testing. Just like response spectra, this is also a convenient way of obtaining the design solution to a structural dynamics problem and is certainly an important tool.