Fuzzy, metrology-sound approach to the identification of sources injecting periodic disturbances in electric networks

The identification of sources injecting periodic disturbances in electric systems is a critical point in the assessment of the electric power quality. The available methods, based on a deterministic approach, do not always provide correct results. Methods based on heuristic approaches, such as those based on a fuzzy inference system (FIS), provide better results but do not allow measurement uncertainty to be evaluated in a straightforward way. This paper applies a modified FIS to the identification of the sources producing periodic distortion in power systems. The method associates an index, provided together with its measurement uncertainty, to each load connected to a point of common coupling capable of assessing whether the load is injecting or suffering distortion and quantifying the severity of the injected or suffered distortion.