HOS and CBR measurement system for PQ assessment

This paper describes an intelligent measurement system for Power Quality (PQ) assessment. Based in Higher-Order Statistics (HOS), its computational intelligence is based in the Case-Based Reasoning (CBR) paradigm. The characterization of the signal is done using the variance, the skewness and the kurtosis over a tuned sliding window. These feature values are introduced in the CBR system and a classification of the signal is performed. The Virtual Instrument (VI) exhibits an accuracy of the 93% over real-life electrical anomalies.