Abstract: With the expansion of the power grid scale and the increase in equipment complexity, traditional power system monitoring faces technical challenges such as insufficient adaptability to static rule libraries, low accuracy in identifying complex events, and lack of precision in assisting decision-making. Based on this, this article proposes a dynamic event based alarm generation and intelligent decision-making system based on the digital station platform. It constructs a dynamic adaptive rule library, multi-source signal fusion reasoning mechanism, equipment health status deep analysis model, and event closed-loop management process, which can accurately identify and efficiently handle power equipment faults in real time. Practical application has shown that the system has significantly improved key indicators such as the accuracy of complex event recognition and the adoption rate of disposal suggestions compared to traditional solutions, providing a technical path for intelligent monitoring of power systems.