Design of Remote Fault Diagnosis System for Hydropower Unit Based on Cloud Platform

As the installed capacity of power plants continues to increase, the complexity of equipment is getting higher and higher, and the importance and difficulty of fault monitoring also increases. In this paper, a human-machine closed-loop remote fault diagnosis system with self-learning function is designed. The establishment of the remote fault diagnosis system for hydropower units based on the cloud platform can fully integrate the accumulated experience and expert knowledge of power plants, and be able to collect and deeply mine the data of related units in different power plants during the operation process, so as to improve the diagnostic efficiency and accuracy of the system. The system introduces the concept of “improvement factor”, realizes the remote interconnection of the inference machine system with domain experts and technicians, and includes human beings in the diagnostic loop to calculate the “improvement factor”, so that the diagnostic process can continuously improve its confidence level. The system consists of three core modules: a fault knowledge base containing confidence factors, a hydropower unit fault diagnosis engine, and an equipment health assessment engine.