The ability of the fast fourier transform to de-noise a strain signal

Abstract

This study aims to develop an algorithm for filtering the noise in the fast Fourier transform. It requires the determination of the cycles, based on the rain flow counting method. This design was further applied to remove the lower amplitude cycles in a 60-second strain signal. Therefore, the results indicated that the method was able to remove more than 10% of the lower amplitude cycles at a frequency of 45 Hz. The filtered strain signal maintained the fatigue damage by more than 90% and further upheld its original characteristics. This study concluded that the developed algorithm was able to identify and remove the noise contained in the strain signal.