The establishment of a set of perfect equipment abnormality early warning and emergency response mechanism, equipment abnormality early warning mechanism can cooperate with the emergency response mechanism, and jointly solve the abnormality, is an important link to provide high quality and stable power supply service. This paper deeply researches the two perspectives of equipment abnormality warning and emergency response mechanism, for the equipment abnormality warning mechanism, this paper deeply explores the digital twin technology and data preprocessing, and combines it with the discriminator in the abnormality detection model to construct a set of equipment abnormality detection and warning model. Experimental results show that the false alarm rate of this set of models is 0.75%, with less error, which verifies the accurate prediction and real-time alarm effect of the model in this paper. For the emergency response mechanism after getting the abnormal warning information, this paper combs through the specific process of the mechanism, analyzes the performance of the mechanism in actual cases, and finds that it has efficient emergency response and disposal process.