华梵前山校区风速统计分布之探讨

Analysis of Statistical Distribution to Wind Speed Data at Huafan Campus

研究针对华梵前山气象站2002年5月至2006年12月之小时风速资料进行统计分析,探讨华梵前山风速资料之分布是否合乎韦伯分布(Weibull distribution)及莱利分布(Rayleigh distribution)。研究中韦伯分布使用动差法(Method of moment)及最大概似法(Method of maximumlikelihood)求取参数;莱利分布使用动差法求取参数。进一步以K-S检定,判断各理论累积机率分布与观测值之累积机率是否相符;再以误差绝对值之总和判断何种分布较为适合。分析结果显示:韦伯分布与莱利分布均可适用于描述全部数据之机率分布。将资料分为夏半年(4-9月)与冬半年(10月至翌年3月)时,韦伯分布为最适用之机率分布函数;且夏半年之参数以最大概似法求得最佳,而冬半年则以动差法较佳。此结果可作为校区风蚀、防灾、生态、环境风场特性及评估风能潜势研究之参考。

This study investigates the statistic distribution of hourly mean wind speed data recorded at west campus of Huafan University from 2002 to 2006. Weibull distribution and Rayleigh distribution are adopted to fit wind speed data. The parameters of these distribution were estimated by using method of moment and method of maximum likelihood. The results show that the statistical distribution of all hourly mean wind speed data is suitably described by the Weibull distribution and Rayleigh distribution. Furthermore, wind speed data are separated into summer parts （April to September） and winter parts （October to March）. The data of two parts are more suitable described by the Weibull distribution. The parameters estimated by method of maximum likelihood are better for summer parts, and the parameters estimated by method of moment are better for winter parts.