Feature Extraction and Classification of Power Quality Events Based on Fast Fourier Transformation and Artificial Neural Network

This paper presents an effective method for detection and classification of Power Quality Events (PQE), based on Fast Fourier Transformation (FFT) event identification Artificial Neural Network (ANN) technique classifying these events. Firstly, synthetic data such as pure sine a reference, voltage sag, swell, flicker, transient, with harmonics are created in MATLAB TS EN 50160 standard. Database 480 PQE waveforms is generated 80 samples each the 6 types waveform randomly different amplitude, beginning occurrence time, time duration, frequency component angle according to type event. FFT used extract features events by decomposing signal. Then, 16384×480 reduced 480×480 applying Principal Component Analysis (PCA) that prevent over-learning, obtain less runtime using computing power reduce storage space. Finally, total classified ANN. 336 training cluster, 72 verification remaining testing. ANN has been trained correctly. The performance examined inserting test into 99.8% PQE. purpose this research provide artificial intelligence assistant can fast accurately advise system operators networks, results also show goal achieved.