Purpose/Aim
Remaining useful life (RUL) estimation of power transformer can effectively guarantee the stability of power grid operation and reduce the cost of equipment maintenance. A data-driven method, Wiener process is proposed for the insulation system, and Weibull distribution is adopted for other key components in this paper. Then, the system-level reliability model is established for the RUL calculation for oil-immersed power transformers.
Experimental/Modeling methods
For insulation system, dissolved gas in oil and degree of polymerization (DP) of insulation paper are employed for RUL estimation. The degradation state of insulation system is obtained by information fusion of historical data. Then, Wiener process is applied for RUL calculation. Finally, with Weibull distribution of other components like bushing, tap-changer, etc., the RUL distribution of the whole transformer is analyzed by series and parallel reliability model.
Results/discussion
The degradation data of insulation system is divided into two parts: training set and test set. The model is trained on training set, and then the trained model is applied for estimation the rest data. The comparison shows an error at about 30% between estimated data and test set.
Conclusions
This paper proposed a data-driven method for RUL estimation of insulation system, and a reliability model for power transformer considering key components. With these methods, the RUL of power transformer can be evaluated individually based on historical data, thus providing effective reference for maintenance strategy.
 

power transformer,life estimation,Wiener process,Reliability Analysis