APPLICATION OF AHP-ENTROPY METHOD AND LSSVM IN THE ASSESSMENT AND FORECASTING OF TRANSFORMER OPERATING STATUS

Niu Guocheng, Hu Zhen, Hu Dongmei

ABSTRACT:In order to meet the needs of fault troubleshooting, maintenance decision and on-line prediction, the method of the transformer health index based on AHP-entropy is proposed. The cross complex matter elements are established between the transformer health index and dissolved gases and typical faults. The subjective and objective weights affecting the transformer health index are determined by Analytic Hierarchy Process (AHP) and information entropy respectively, and the transformer health index is calculated quantitatively by matter element-maximum information entropy, the calculation is used as the predicting information, and the Least Squares Support Vector Machine (LSSVM) is used to predict the transformer operation status. the scheme provides a new feasible method to evaluate, quantify and forecast the transformer operation status.

Keywords:Transfomer, Health Index, Analytic Hierarchy Process (Ahp), Complex Matter Element, Support Vector Regression (Lssvm).