A new approach using Least Squares Support Vector Machines (LS-SVM) to predict Furan in power transformers

摘要

W artykule opisano metodologię prognozowania obecności Furanu w transformatorach energetycznych. Na wejściu podawane są takie parametry jak ilość rozpuszczonych gazów CO i CO_2. Metoda opiera się na wykorzystaniu algorytmów LS-SVM.%LS-SVM present recently more efficiency in different industrial applications like medicine, engineering and power systems. This paper describes a methodology that was developed for the prediction of Furan in power transformers. The methodology uses as input variables such as the dissolved gases (CO and CO_2). The approach presents the advantage that can reduce the time vs. laboratory tests. The validity of the approach was examined by testing several power transformers. LS-SVM gives a good estimation of results which are validated by experimental tests.