Distribution network transformers play an important role in the reliable and economical operation of the power system.In the past

maintenance personnel spent a lot of time checking the distribution transformers one by one.Nowadays

with the widespread deployment of electrical energy data collection systems for power users

useful measurement data can be provided and used for further condition monitoring.Therefore

this paper proposes a datadriven abnormal state monitoring data collection algorithm for distribution network transformers

which can send abnormal state alarms to operators and maintenance staff in a timely manner.In the proposed algorithm

the Spearman rank correlation coefficient is used to display the correlation between the phase currents

and its t-statistic is used to determine whether there is an abnormality in the data collection based on the hypothesis test.Finally

the actual collected data of Zhejiang Power Grid is used to verify the effectiveness of the proposed algorithm

and the characteristics of normal and abnormal situations are analyzed separately.Sensitivity analysis is performed for different significance levels and sampling rates to consider its impact on the monitoring results;the application in actual power systems is also given to prove the practicability of the algorithm.