Abstract: In the big data environment, the power network monitoring data flow is large and changes quickly, the reliability of the abnormal link judgment of the power grid communication is poor, and the sequence anomaly detection results are distorted. This paper proposes a big data security analysis based on the power network monitoring data research on key technologies of the platform. Through the design of the power network communication abnormal link judgment method, the multi-dimensional entropy sequence anomaly detection method design analyzes the key technologies of the platform.The multi-dimensional entropy sequence anomaly detection method is used to measure the distribution characteristics of the network flow data collected on the abnormal link, and the entropy value sequence of the power network flow monitoring data in each dimension is obtained, and the support vector machine is used to measure the network flow data in each dimension.