风速时间序列复杂度分析

Complexity analysis of wind speed time series

为研究测风塔不同高度以及气象站实测逐时风速的复杂度,文中以山东一风电场为例,收集测风塔5个高度层及气象站同期的1年逐时风速数据,采用广义极值分布理论以及多尺度熵模型对不同高度不同时间长度序列的风速时间序列进行复杂度分析。结果表明,测风塔除10m高度外的风速数据更符合Frechet分布;而10m高度和气象站风速数据更符合Weibull分布;随着时间尺度的增大,不同高度风速数据的复杂度首先变高,在8~12h处样本熵达到最大值,然后复杂度逐渐降低。

In order to study the complexity of the measured hourly wind speed at different heights of a wind tower and the local weather station,the wind speed data of the tower and the station for one year in Shandong was collected.The generalized extreme value distribution theory and the multi-scale entropy model were used to analyze the complexity of the wind speed at different heights and different time scales.The results show that the wind speed data of the wind tower is more in accordance with Frechet distribution except that at the height of 10 meter which is more in accordance with Weibull distribution as well as the meteorological station.As the time scale increasing,the complexity of the wind speed data increases firstly,and then decreases gradually,the sample entropy reaches maximum in the range between 8h and 12h.