Data center interconnection and relay station full life cycle monitoring method and system

Abstract

The invention relates to the technical field of data center interconnection and relay station full life cycle monitoring, and discloses a data center interconnection and relay station full life cycle monitoring method and system, and the method comprises the steps: constructing an electrical disturbance reference; performing causal effective interval screening by adopting a link health state reverse constraint mechanism; and finally predicting delay jitter and error rate through the time sequence causal convolutional network. And performing real-time early warning monitoring based on the predicted delay jitter and the bit error rate. Compared with a link monitoring method based on traditional threshold alarm in the prior art, the technical problem that the potential causal relationship between current fluctuation and link performance degradation cannot be effectively captured in the prior art especially in the face of complex power system fluctuation is solved. According to the method, link prediction and risk assessment based on historical electrical disturbance are realized through time sequence causal modeling and self-adaptive monitoring, so that a traditional scheme which only depends on a static threshold value and data presetting is avoided, and the early warning capability of link monitoring is improved.

Images