Abstract: With the continuous increase in electricity demand, the risk of faults in ultra-high voltage power grid equipment has significantly increased, and there is an urgent need for effective warning systems to ensure the stable operation of the power system. This study takes a certain ultra-high voltage power grid in Henan Province as an example, and designs and deploys an intelligent early warning system based on big data analysis and machine learning models. The system achieved an accuracy of 96.8% in equipment fault prediction by collecting multiple data such as temperature and humidity, voltage, and current, and applying a fusion deep learning model of LSTM and CNN, significantly reducing equipment failure rates and maintenance costs. The results show that the system exhibits significant advantages in fault prediction, equipment failure rate, and maintenance costs, significantly improving the operational stability and power supply reliability of the power grid, and providing a new solution for fault warning of ultra-high voltage power grid equipment.