In order to facilitate the dynamic monitoring of the safety of equipment

we proposed a method for real-time assessment of power transformers by fusing different data of equipment index in update periods. Based on the construction of the evaluation index system

two kinds of score functions on data were proposed for real-time monitoring indices and preventive test indices

respectively. At the same time

based on the equipment samples and the assessment time of the transformer

a method for calculating the index weights combining Bayesian posterior probability and time-varying function was proposed. To solve the imbalance problem of the comprehensive indices

we introduced the variable weight formula to adjust and calculate the weight coefficients. Sample calculation results show that the correct rate of evaluation is up to 93.5%. By making full use of the data of the equipment itself

combined with mining the statistical laws from the same type of equipment

the method is suitable for real-time condition assessment of power transformers.