摘要
为了保障风电机组的安全经济运行,需要对风机关键部位的实时载荷进行监测和预测。因此,提出了一种基于Informer模型的风机叶根实时载荷预测方法。首先使用OpenFAST软件对风机进行仿真计算,然后利用仿真数据搭建Informer预测模型对风机叶根实时载荷进行预测,最后给出了模型预测数据与实际数据的曲线并对预测结果进行了评估,证明了该方法的可行性。通过与LSTM和GRU两种常规模型的预测结果进行对比,证明了Informer模型有更优的预测性能。
In order to ensure the safe and economical operation of wind turbines,it is necessary to monitor and forecast the real-time load of key parts of wind turbines.Therefore,a method for blade root real-time load prediction of wind turbines based on the Informer model is proposed in this paper.Firstly,the OpenFAST software is used to simulate the operation of wind turbine,then an Informer prediction model is built with the simulation data to predict the blade root real-time load of wind turbine,and finally,the curves of model prediction data and the actual data are given,and the feasibility of this method is proved by evaluating the prediction results.It is proved that the Informer model has better prediction performance by comparing with the prediction results of LSTM and GRU models.
作者
王伟
王海云
黄晓芳
WANG Wei;WANG Haiyun;HUANG Xiaofang(College of Electrical Engineering,Xinjiang University,Urumqi 830046,Xinjiang,China;Xinjiang Goldwind Science and Technology Co.,Ltd.,Urumqi 830026,Xinjiang,China)
出处
《水力发电》
CAS
2023年第9期85-89,共5页
Water Power
基金
自治区重点研发计划资助项目(2022B03031)。