544 / 2022-03-29 16:49:41

Remaining Life Prediction of Power Transformers Based on Data Fusion and Wiener Model

power transformers,life prediction,data fusion

Purpose/Aim

Oil-immersed transformers produce gas, furfural, and other related substances during the aging process. These substances help to reflect the operation state of the transformer together with the degree of polymerization (DP). It is proposed to predict transformer life by fusing transformer multi-variate data and to solve the randomness of the parameters by the Wiener model.

Experimental/Modeling methods

Considering the random influence of internal and external working environment, the Wiener model is widely used in degradation modeling due to its strong adaptability. The parameters of the Wiener model can be updated by the Bayesian and maximum expectation algorithm. Then the probability distribution and life expectation of power transformers remaining life are obtained.

Results/discussion

The way of data fusion put an impact on the overall change trend of power transformers and equipment fault in time.

According to the established Wiener model with respect to the working history of ten 500kV power transformers in the field, the remaining life of power transformers is calculated by this method can reflect individual differences of power transformers in different operating environments and operating conditions.

Conclusions

The proposed method of combining multivariate data and Wiener model is feasible to predict the remaining life of transformer in real time.