In order to further optimize the condition maintenance mode for transformers

a condition maintenance mode based on non-blackout detection was proposed in this paper. Firstly

the data of transformer imminence

serious defects and faults of 8 provincial power companies from 2005 to 2014 were collected

and the detection efficiency of state parameters and detection modes was analyzed. Secondly

based on the judgement of expert

the key state parameters were optimized and the blind detection rates were evaluated. Finally

the feasibility of canceling the routine test of power outage was analyzed. The results show that transformers are capable of performing condition-based maintenance based on online detection without power cut. Compared with the traditional periodic maintenance mode

less state parameters are needed for the non-blackout detection proposed in this paper

which can shorten the power outage time and improve the maintenance efficiency.