Journal of Agriculture & Life Sciences

ISSN 2375-4214 (Print), 2375-4222 (Online) DOI: 10.30845/jals

Classification of Power Signals Using PSO based K-Means Algorithm and Fuzzy C Means Algorithm
B. Majhi, S. Sabyasachi, S. Mishra

This paper presents a new clustering technique and pattern classification of power signal disturbances using a modified version of S-Transform, which is obtained by taking the Inverse Fourier transform of S-Transform called as modified time-time transform (TT-transform). The TT-transform is a signal processing technique which is used for visual localization, detection, and power signal disturbance pattern classification. The TT-Transform, a new view of localizing the time features of a time series around a particular point on the time axis. TT-Transform has good ability in gathering frequency; it gathers the high frequency signals in diagonal position of the spectrum and suppressing the low frequency signals. The diagonal of TT-Transform represent a simple frequency filtered version of the original signal. The noise can be separated from the effective signal, which can improve the signal-to-noise ratio by using TT-Transform. The extracted features are fed as the input to a Fuzzy C-Means clustering algorithm (FCA) and K-Means algorithm for power signal disturbance pattern classification. To improve the pattern classification of the fuzzy C-means and k-means algorithm, the cluster centers are updated using particle swarm optimization technique. Comparison of both the algorithm is made for power signal disturbance pattern classification accuracy

Full Text: PDF