Time-Frequency Analysis Of Accelerogram

Abstract

In signal processing, time-frequency analysis consists of studying a signal in time and frequency domains simultaneously. Rather than viewing a 1-dimensional signal (a function, real or complex-valued, whose domain is the real line) and some transform (another function whose domain is the real line, obtained from the original via some transform), time-frequency analysis studies a two-dimensional signal– a function whose domain is the two-dimensional real plane, obtained from the signal via a time-frequency transform. These high-level representations such as time-frequency maps convey a wealth of useful information, but they involve a large number of parameters that make statistical investigations of many signals difficult at present. In this paper, we will describe a method that performs a drastic reduction in the complexity of time-frequency representations through modeling of the maps by elementary functions, Artificial Intelligence, and Machine learning. The method is validated on artificial signals and subsequently applied to signals recorded at original stations. We will show different methods of doing Time-frequency analysis using techniques like FFT(Fast Fourier Transform), wavelet methods, and how by applying Artificial neural networks, deep learning can significantly reduce the complexity of time-frequency analysis with more return in result. We will try to validate the advanced technological improvement in this field to show the potential and promise of technology in this area.