摘要
通过神经网络建立数学模型,利用 Multibrid M5000风力机实测的运行信号对叶片拍打弯矩、叶片弦向弯矩、塔筒弯矩和塔筒转矩的等效载荷进行了评估。同时分析了不同运行参数对评估效果的影响程度。通过与实测结果比较表明,该方法可以在避免测试风力机载荷的情况下获得载荷的评估结果,具有较高的精度,可以用来预测风力机的寿命。
Load test for wind turbines is complicated and expensive work. Equivalent load evaluation method of wind turbine through neural network was presented in the paper. The statistic data of running historical signals, which were aquired from wind turbine Muhibrid M5000, were used as input data of the neural network model. The influence degrees of different nmning historical signals on estimation results were analysed to aquire important input parameters of neural network. Flapwise and edgewise bending of rotor blade, tower top tilt moment and tower top roll moment were estimated. In comparison with calculation results based on tested load data of Muhibrid M5000, the estimation results of wind turbine loads represent very high precision and can be used to predict the life of wind turbine.
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2008年第12期1456-1459,共4页
Acta Energiae Solaris Sinica
关键词
等效载荷
神经网络
风力机
疲劳
equivalent load
neural network
wind turbine
fatigue
