中文摘要：

针对兆瓦级风力机部件承载的载荷范围宽、部件寿命时间长所导致的短时测试或仿真所采集的数据不完整的问题,采用改进的统计外推法计算疲劳损伤密度,选定拟合概率分布的阈值,消除大量小循环使载荷分布尾部失真的影响;对大于阈值的数据进行概率分布拟合,建立一种更接近载荷分布尾部的概率分布模型,实现风力机疲劳载荷分析。仿真计算表明,与传统的统计外推方法相比,所采用的改进统计外推法在实际应用中准确性更高,能够更好地适用于风力机部件疲劳分析。

英文摘要：

Because of wide load range and long service life of mega watt scale wind turbine components, there is a mass of data among them, however incomplete data of measurement or simulation only covers part of those. In order to solve this problem, a method of improved statistical extrapolation is proposed to compute fatigue damage density. A suitable threshold of probability distribution is chosen to remove the effect that large number of small cycles distort the tail of distribution;The data above the threshold is used for fitting of probability distribution to establish a model that closes to tail of the distribution, which can realize fatigue load analysis of wind turbine finally. Simulation results indicate that comparing with the conventional approach, the iteration method of improved statistical extrapolation using in this paper has more precise in practical application and can be adopted to fatigue analysis of wind turbine components.