摘要
Abstract
Hydropower generation is a clean energy technology that converts the energy of water flow into electrical energy.The stability and safety of the operation of the pump set equipment in hydropower stations are of vital importance.The study established a centrifugal pump bearing test bench,collected data using vibration sensors,and used a CNN model for feature extraction and fault diagnosis,achieving real-time monitoring of the operating status of pump equipment.The experimental results show that the accuracy,recall,and precision of the CNN model for bearing fault diagnosis on the measured set are 0.961,0.956,and 0.972,respectively,which are superior to SVM and decision tree models,demonstrating strong fault diagnosis capabilities.In the case of scarce training data(200 sets),the accuracy of the CNN model is still greater than 0.85.关键词
水电站/泵站运行/监测分析Key words
hydropower station/pumping station operation/monitoring analysis分类
建筑与水利
