Research of harmonics in power system signal using gaussian’s distribution overlapping by receiver operating characteristics (Roc) curve

Abstract

Harmonic analysis of the power system signal is a proliferating research in the field of electronics technology. Whenever, we analyze odd and ever harmonics are present in the signal, imperative operation is needed to transform from the time domain to the frequency domain. Hence, all the researchers are utilizing the Fourier Transform technique is very effective for the analysis of odd and even harmonics in the frequency domain. In the past two decades, Wavelet Transform is a wonderful technique to analyze the harmonics both frequency and time domain as well. The analysis of harmonic and its probability distribution are most important for the purpose to predict the harmonic effects in the present situation. We treated all the harmonics and its corresponding frequency distribution are considered as a zero mean unit variance. The overlapping these distributions (small, medium, large) are analyzed with help of statistical data processing technique. It is one of the most important basic plots in the decision theory and it provides the constructive decision about the overlapping of a frequency distribution in power system signal. The curvature as a plot of sensitivity and specificity underlying the harmonics are present and not present distributive (Gaussian). The above determined values are lying in the interval probability [0, 1] and it is depends only the nature of the dataset. In this paper, we explained with help of MATLAB and level of understand the basic concept of ROC is demonstrated. The dataset is drawn from the example of odd and even harmonics are generated and the probability distribution as input to our MATLAB program.