摘要
风力发电技术在全球实现规模化发展,风电机组正在朝着大型化、智能化与数字化方向发展。这里利用深度学习中的卷积神经网络技术建立了大型风电机组载荷预测模型,该预测模型可以通过对风电机组运行数据和主要部件如塔架、机舱以及叶片载荷数据进行学习与训练进而对风电机组关键载荷进行预测与评估。应用建立的载荷预测模型对某型3.XMW风电机组DLC1.2所有设计工况载荷进行了预测。结果表明,预测模型对时域载荷、频域捕捉以及等效疲劳载荷的预测都能很好的跟随测试数据,精确度均达到90%以上,满足工程实际的要求。
Wind power generation technology has achieved large-scale development in the world,and wind turbine is developing towards large-scale,intelligent and digital.A load prediction model for large-scale wind turbines is established using convolu-tional neural network technology in deep learning.The key loads of wind turbines are evaluated and predicted through learning the operation data and load data of main components such as tower,nacelle and blade by the prediction model.The design loads of DLC1.2 for a 3.XMW wind turbine are calculated by the prediction model.The results show that the test data are followed well by the predicted data for time domain,frequency domain and equivalent fatigue loads.The accuracy reaches more than 90%,which meets the requirements of engineering practice.
作者
尹尧杰
褚景春
姜培学
YIN Yaojie;CHU Jingchun;JIANG Peixue(Guodian United Power Technology Co.,Ltd.,Beijing 100039,China;CHN Energy New Energy Technology Research Institute Co.,Ltd.,Beijing 102209,China;Dept.of Thermal Engineering,Tsinghua University,Beijing 100084,China)
出处
《机械设计与制造》
北大核心
2025年第6期217-221,共5页
Machinery Design & Manufacture
基金
国家重点研发计划(2020YFN1506700)大型柔性叶片气动弹性设计关键技术。
关键词
风电机组
数字化
卷积神经网络
载荷预测
时域载荷
疲劳载荷
Wind Turbine
Digital
Convolutional Neural Network
Load Prediction
Time Domain Load
Fatigue Load
