Abstract:Hydro generator is the key equipment of hydropower station, its running state is directly related to the safety production of hydropower station. At present, in order to master the running status of hydro generator, we mostly by analysis of the running data of the state monitoring system and scheduled maintenance check. Usually, faults can only be found after they have occurred, and the running state of equipment cannot be predicted in advance. This paper introduces a fault diagnosis system of hydro generator based on big data platform. The system consists of data processing platform, model algorithm platform and visual display platform, is a typical application based on big data platforms and Internet technology. It mainly excavates the real-time and historical data of the existing information systems such as the computer monitoring system and the state monitoring system, and then extracts the rich value information contained in these data through a series of association algorithms to realize the running status monitoring, health evaluation, trend warning, fault diagnosis and maintenance guidance of the unit.