Flicker’s sources identification using a case-based reasoning prototype

Abstract

Flicker is an issue of power quality that affects both the utility companies and consumers. To mitigate and eliminate flicker, the first stage is to identify the source. This paper presents an application of artificial intelligence techniques, namely a knowledge-based system, dedicated for detecting the flicker source. The application is a prototype that was built using case-based reasoning. The main elements of the prototype are the cases that were obtained from in-situ measurements at a power substation over an observation period of a year. The prototype was tested using data from a different power substation. The results show that the prototype detects properly the sources, confirming the utility of artificial intelligence in this field ofpower quality.