Development of remaining life assessment for oil-immersed transformer using structured neural networks

Remaining Life of the oil-immersed transformer is decided due to deterioration of the winding insulation paper. The furfural method is conventionally used to estimate the remaining life. However, the results are obtained as wide ranges between upper and lower limits. Therefore, a more accurate estimation method has been expected. This paper proposes the remaining life assessment for oil-immersed transformer using structured neural networks and ensemble technique. The authors have estimated the remaining life using proposed method for 300 transformers or more. As a result, appropriate replacement time of transformer and appropriate maintenance scenario can be planned.