A Method of Abnormal Detection for Power Grid Control System Metrics Based on AI Platform

In the intelligent maintenance of power grid control system, a real-time anomaly detection algorithm for metrics based on sliding windows is proposed. Then, short-term prediction and dynamic threshold are distributed constructed based on AI platform. Finally, multi-dimensional correlation analysis are achieved. Based on the metric of a system, a case study was conducted to verify that the method proposed in this paper is significantly superior of timeliness, accuracy and efficient.