Comparative study of 1D de-noising techniques using induction motor fault signals

Abstract

Removing noise from the original signals has always been challenging. A lot of denoising techniques and algorithms has already been published and are being used now-a-days. But knowing which de-noising method is better we need to compare the performance of their applications so that a new group of people can start working instantly knowing which method is better than the other considering specific parameters. In this thesis, we will compare the de-noising techniques using discrete wavelet transform (DWT), empirical mode decomposition (EMD), Gabor filter, butter filter, low pass filter, high pass filter, band stop filter, Hilbert filter, Median filter and Q function. To evaluate the performance of the state-of-art models we will utilize an induction motor dataset that consist of inner, outer, and roller fault signals including healthy/normal signal. Finally, the performance will be measured using the following parameters: signal to noise ratio (SNR), peak signal to noise ratio (PSNR), mean square error (MSE), root mean square error (RMSE) and mean absolute error (MAE). We will also find the structural similarity(SSIM), structural dissimilarity(DSSIM) and cross correlation (CC) between the parameter.