Maximum Correntropy Criterionの電力系統問題への適用

抄録

Various practical problems including power system problems have been solved using least squares methods as basic techniques. However, the methods cannot be applied when outliers exist in data or errors are not normally distributed. In order to tackle the challenge, various methods have been developed. Recently, maximum correntropy criterion (MCC) has been proposed in the signal processing field and it has been verified to be effectively applicable to various problems including power system problems. This paper briefly overviews the MCC and summarizes MCC applications to power system problems. Applications to distribution system state estimation and connection phase estimation of pole mounted distribution transformers are shown as examples.