In order to improve the management of distribution system and the accuracy of optimization strategy

it is necessary to correct the capacity of distribution transformers. Moreover

it is time-consuming and laboursome to correct the distribution systems transformer capacity in artificial way. Hence

based on the analysis of the on-line operation data

an on-line data-driven model

which is used to detect the capacity of distribution transformer

was proposed in this paper. First

a high precision chain evaluation method was proposed to resolve the model

where the threaded binary tree was utilized to establish the topology relationship of distribution transformer group

and the distribution transformer with different joint groups was distinguished by correlation coefficient method. Then

the short circuit impedance of Dyn11 distribution transformer and the zero sequence impedance of Yyn0 distribution transformer were calculated and compared with the standard reference value. Therefore

the online evaluation of the distribution transformer capacity was realized. The results of test examples show that

when adopting the proposed method the correction rate of capacity correction of distribution transformers in the same feeder with different connection groups can reach 100%; the proposed method can effectively detect the transformer capacity of the distribution systems

and it can be applied to the large-scale on-line transformers in the distribution systems. The research results can provide a reference for the capacity correcting of distribution transformers.