Publication

Abstract : The extensive use of non linear loads results in proliferation of voltage and current harmonics in an interconnected power system. Harmonic estimation without the information on the topology can be done using Independent Component Analysis(ICA). In this work, the time structured ICA algorithms - Fast ICA (FICA), Joint Approximate Diagonalisation of Eigen Matrices (JADE) and Entropy Bound Minimisation (EBM) are employed for harmonic voltage and current estimation by applying the principle of superposition. These algorithms are implemented on a IEEE 14 bus system. The Spearman's coefficient and the concurrent deviation method of correlation coefficient are formulated for performance comparison. However, when FICA is integrated with weight adjusted second order blind identification technique(WASOBI) resulting in FIWAS algorithm superior performance for estimating the even harmonics and other conventional harmonic frequencies is observed