An ICA-based method for adaptive decomposition of multiple PQ disturbances

This paper proposes a method for adaptive decomposition of Multiple Power Quality Disturbances based on Independent Component Analysis. This method aims to decompose power system signals (voltage or current) into their isolated disturbances when a multiple disturbance situation occurs, providing more specific information about the different disturbances which occur simultaneously. This work proposes a methodology for the Independent Component Analysis application when only a single measured signal is available (single channel). Also, it presents a procedure based on the Least Mean Square algorithm to recover the scales of the independent component estimations. Besides this, it shows that the proposed method is an adaptive filter bank that blindly separate out spectrally distinct independent disturbances.