Application of algorithms and artificial-intelligence approach for locating multiple harmonics in distribution systems

A new method is proposed for locating multiple harmonic sources in distribu tion systems. The proposed method first determines the proper locations for metering measurement using fuzzy clustering. Next, an artificial neural ne twork based on the back-propagation approach is used to identify the most l ikely location for multiple harmonic sources. A set of systematic algorithm ic steps is developed until all harmonic locations are identified. The simu lation results for an 18-busbar system show that the proposed method is ver y efficient in locating the multiple harmonics in a distribution system.