Arc fault detection in a power system network is a vital diagnostic test for condition monitoring in order to remain the consistency of the service. The issue of electrical faults has been largely investigated in the literature to foresee the existence of arc fault at the earliest stage. However, the identification measurement has faced noise disturbance from surrounding area. Thus, this paper presents an arc fault signals detection and analysis using Discrete Wavelet Transform (DWT) technique as a denoising technique. The effectiveness that influence the different types of mother wavelets is analyzed based on calculated signal to noise ratio (SNR) before denoising and after denoising, mean square error (MSE), correlation coefficient (CC) and mean absolute percentage error (MAPE). Each mother wavelet will be used to extract important features of a voltage signal from a single measurement point that were performed under 4th decomposition level using Universal (Sqtwolog) thresholding rule with soft thresholding function. The results are manipulated based on Haar (haar), Daubechies (db), Symlets (sym), Coiflets (coif), BiorSplines (bior) and Reverse-Bior (rbio) mother wavelets. The outcome demonstrates the data analysis approaches can provide better understanding to obtain the most optimum mother wavelet hence can enhance the performance of arc fault detection using wavelet transform as a denoising purposed.
