Health assessment model of power transformer based on dissolved gas analysis by support vector machine

Power transformer in grid is of vital importance to guarantee safe operation of the power system. Nowadays dissolved gas analysis (DGA) is recognized as an effective method for detecting the initial fault of power transformer. Support vector machine (SVM) with Structure Risk Minimization (SRM) is powerful for the problem with small sampling, nonlinear and high dimension, which is based on mathematics theory. This paper establishes health assessment model based on DGA by using SVM. Health estimation is used to deal with the initial fault timely and prevent fault diffusion and dissemination. Then we choose appropriate kernel function and corresponding parameter to ensure the availability. Finally, a case is studied. The result indicates that the model can effectively estimate the health state of power transformer.