LAUSR

Abstract Based on the association rules and variable weight synthesizing theory, a new insulation condition assessment (ICA) method for power transformers is presented, employing fused information in both time and space dimensions. With the uncertainty and fuzziness of available data from the power transformer concerned taken into account, a multilateral transformer ICA model based on fuzzy and evidence reasoning is developed. The ICA model contains three layers: target layer, factor layer, and index layer. The association rules and variable weight synthesizing theory are applied to determine the variable weight coefficients of factors and its indicators in the assessment model. A fuzzy membership function is formulated to describe the factor layer on assessment model. The evidence reasoning and the multilateral assessment scheme are adopted to merge the information of each factor in time and space dimensions, and then the assessing result of the transformer is obtained. Numerical results based on practical scenarios demonstrate that the proposed method is feasible and efficient.