With the rapid development of energy Internet, power Internet of Things and other technologies, power information and communication technology has become an indispensable part of the power grid. However, due to the frequent occurrence of various types of network attacks, the security and stability of the power information communication system are greatly threatened. Artificial intelligence, with its powerful learning and decision-making abilities, has become an important direction in network security research. This article takes power system security as the research object and production based artificial intelligence as the research object, studying an adaptive model framework applicable to power grid security issues. A complete power grid situational awareness system has been constructed from three levels: collection layer, storage layer, analysis layer, and response layer. And combined with practical cases, conduct experimental research on the proposed theories and methods to provide innovative solutions for solving power grid safety issues.