The work aims to monitor the status of complex equipment and diagnose its faults more effectively and accurately. The fundamental principle, characteristics, limitations and research status of important methods in fault diagnosis technologies in recent years were summarized. Based on a large amount of literature, the development trends of modern fault diagnosis technology were discussed based on computer technology, signal processing technology, artificial intelligence technology and Internet technology. The fault diagnosis technology mainly researched the reflection of the change of machine or unit operating status in the diagnostic information. It was divided into model-based, signal-based and artificial intelligence-based technology. With the continuous development and cross infiltration of basic disciplines and cutting-edge disciplines, the fault diagnosis technology is also constantly innovating, and its future development trends are mainly focused on combining, integrating or fusing different artificial intelligence technologies in some way and the open remote collaboration diagnosis technology.