Shaft Run-out Trend Prediction of Water Turbine Generators and Fault Identification of Hydroelectric Units Based on Xgboost Algorithm

Hydropower has become an important weight for stable operation of the power grid due to fast start-up of units and rapid load adjustment. In order to better support the power grid and guarantee the safe operation of hydroelectric units, this paper proposes a method based on XGBoost algorithm to predict the shaft run-out trend of water turbine generators and identify the faults of hydropower sets. Firstly, source data from sensors and deployed systems within the hydropower station are collected through intelligent terminals. Then, the source data is transmitted to the data storage and processing module through the communication gateway. After data preprocessing, all the data is classified and stored according to the unified standard. Finally, according to expert experience, representative features of data will be input into XGBoost algorithm for regression and classification model training, so as to realize shaft run-out trend prediction and fault identification. Through actual deployment and verification, the proposed method can improve the intelligence and operation optimization level of the hydropower station, as well as effectively identify the potential equipment hazards, thereby ensuring its safety and stability.