Abstract: Combined with the current application of online monitoring system in hydroelectric generator sets and the large number of trend faults in water turbine faults, and in order to identify the trend faults more timely and effectively, the trend anomalies in each condition monitoring amount during the unit operation are summarized and classified in this article. At the same time, the effective characteristic parameters of the trend abnormal fault identification are extracted, and the different types of trend anomalies are evaluated by automatically calculating the values of these characteristic parameters under the same working condition, so as to realize the early warning and diagnosis of the health status of the unit operation. The application cases of different types of trend anomalies and characteristic parameters in the fault identification are also listed in this article, which would verify the practical application value of the trend fault identification method in the operation and inspection of hydropower units.