The 500-meter spherical radio telescope (FAST) has be fully commissioned in 2019, and the operation and maintenance of FAST has been the most prominent task. Therefore, it is necessary to develop a health management system. This paper uses digital twinning technology to build the real-time virtual model of FAST that can simulate the behavior and operation of actual devices. The model collected various data from FAST during operation, including mechanical vibration, temperature, humidity, etc., for analysis and diagnosis. Machine learning and deep learning are used to process and analyze the data to achieve the monitoring and diagnosis of the health status of FAST. The results confirm that the health management system developed in this paper enables the monitoring and diagnosis of FAST on a real-time basis, predicting and detecting any potential health problems in a timely manner. The system can help the maintenance personnel to find and solve the possible problems of FAST in time and avoid equipment shutdown or failure, thus improving the reliability and stability of FAST.