APPLICATION OF SVM AND PNN NEURAL NETWORKS FOR CLASSIFICATION OF POWER DISTURBANCES IN DISTRIBUTION NETWORKS CONSIDERING VOLTAGE FUCTUATION AS EXAMPLE

摘要

In this paper the problem of classification of different voltage fluctuations is discussed. Firstly the need for automated power quality assessment, especially voltage fluctuation, is shown. The SVM and PNN networks are then introduced, compared and considered as applicable tools for classification issues. The future extraction for neural classifier is done exploiting the space phasor idea. Disturbed signals were generated using a model built in Matlab. Finally, the results of classification of disturbed signals are presented.