Transients detection using Artificial Neural Network based HS-transform

Abstract

This paper proposes HS-transform technique and Artificial Neural Network (ANN) algorithm for the detection and classification of voltage transients. Transients due to energization of capacitor banks has been considered for the analysis. In the proposed methodology, the HS-transform is employed to generate the instantaneous frequency vectors for S-matrix of the simulated transient signals. Using the statistical features of S-matrix vectors, the occurrence of transient is easily distinguished from the normal condition and from the other disturbance conditions by employing ANN algorithm. The advantage of the proposed algorithm is its accuracy in identifying the presence of transients and its classification accuracy. The performance of the proposed algorithm is manifested by the simulation of oscillatory transient event using MATLAB/SIMULINK software. The test results proved that the proposed algorithm is simple, fast and more accurate in identifying the event.