中文摘要：

利用风电功率波动数据,使用滑动平均法、概率分布理论、1阶差分法等,分别建立了最佳概率分布模型、min级分量模型和变化速率模型,利用Matlab软件编程,得到了风电功率的波动特性的最佳概率分布为t location-scale分布并且得出了以较长时间为时间窗衡量风电功率的波动会损失短时风电功率波动的信息.

英文摘要：

According to the data of wind power, moving average method, probability distribution theory and first difference method are used to establish best probability distribution model,minutes level component model and change rate model. Trough Matlab, the optimal probability density function-t location-scale of wind power variations and the massage which lost within different time widows has been found.