With the continuous development of China’s smart grid and power Internet of Things, China has formed a large amount of big data. These big data have the characteristics of large quantity, diverse types, high value, and fast speed, laying a solid foundation for promoting large-scale data applications. On the power generation side, large-scale grid integration of new energy sources such as wind and solar energy has broken the traditional relative static state, making the measurement and management of electricity usage more complex. Secondly, due to the inability to store electrical energy, the safety situation in the power industry is very complex. On the power generation side, with the continuous evolution of the new generation power grid, the grid supply chain based on high elasticity and big data will gradually be replaced. This article is based on the above issues, exploring the big data analysis and smart grid security event monitoring and response in the power Internet of Things, and constructing a mathematical model through Bayesian network algorithm for power grid security event monitoring and early warning. The experimental results show that the Bayesian network model has a false alarm rate of 0.03 at low loads, which is lower than other methods.