TT-ACO based power signal classifier

This paper intends to propose a novel clustering method based on ant colony (AC) algorithm. A new approach called TT-transform based time frequency analysis is used in processing the non-stationary power signal disturbances. The time-time transform is the inverse Fourier transform of S-transform. The proposed model is demonstrated using feature vector from the domain of power signal analysis, yielding promising results. Visual localization, detection and classification of non-stationary power signals problem is carried out through TT-transform to generate time-frequency contours for extracting relevant features and certain pertinent feature vectors are applied to the fuzzy C-means algorithm with ant colony optimization for power signal classification. From simulation results, it is shown that the proposed algorithm has superior performance when compared to particle swarm algorithm.