Feature Characterization of Power Quality Events According to Their Underlying Causes

In this paper are discussed features resulting from the analysis of each underlying cause of power quality (PQ) disturbances. A feature set for each underlying cause is analyzed; likewise, another feature set able to distinguish between the different underlying causes is also analyzed. The proposed feature sets are useful for building suitable frameworks for automatic identification of the underlying cause of events stored in PQ databases. The features are analyzed using real-world and synthetic voltage a current waveforms. A rule-based framework is also proposed and tested. The rules are based on the analyzed features. The framework has obtained a good classification rate of 95.9% demonstrating a high accuracy distinguishing between the different underlying causes in PQ events