Application of Improved Independent Component Analysis Algorithm to Electric Power Load Profile Estimation
Qingquan Sun · Gao dianya jishu · 2010
FastICA(Independent Component Analysis) algorithm is applied to estimate active load profiles without prior knowledge of the electric network model parameters or topology. However,estimation accuracy can be lowered considering the measurement noise because the conventional algorithm has to separate the observed data including the noise. Consequently,we improved the FastICA to meet the demands of more accurate results depending on the induction of noise during the process of whitening. The experimental results based on IEEE 14 system show that estimation accuracy of the improved FastICA is higher than that of the basic FastICA. It is concluded that the improved FastICA is appropriate for the load profiles estimation with noise.