An integrated three-level transformer condition assessment model based on optimal weights and uncertainty theory

This paper proposes a three-level transformer condition assessment model based on matter-element cloud theory and evidential reasoning decision-making theory. Both quantitative and qualitative factors are considered in the model. Matter-element cloud model is utilized to integrate the quantitative indices with their optimal weights to assess the condition of items, including Dissolved Gas Analysis(DGA), electrical testing and oil testing in the first level in the form of association degrees with state grades to generate the original basic probability assignments for the second-level model. D-S evidential reasoning decision-making model is utilized in the second level to assess the condition of systems including quantitative system and qualitative system and in the third level to assess condition of the overall transformer. Both fuzziness and randomness are taken into consideration in this model. Experimental cases confirm that the assessing model is capable of offering an overall assessment of the observed transformer condition and assisting the maintenance strategy development.