The information system is a key part of the construction of the power Internet of Things. Once the power information system is in danger, it will affect the safe and stable operation of the entire power grid. The current security protection strategies based on power information systems mainly focus on traditional protection and detection methods. However, many threats occur in a short period of time and are easily overlooked and cannot be detected in time. These threats usually have a huge impact on the power information system and interfere with its normal operation. In response to this problem, this paper proposes a power information system security status monitoring mechanism based on big data analysis. It will be based on fuzzy clustering to effectively evaluate the network operation. At the same time, a network situation awareness model that combines game theory and machine learning is proposed to effectively reduce the security operation and maintenance risks of power information systems. The simulation results show the effectiveness and practicability of the security state analysis strategy proposed in this paper.