气象再分析资料在复杂山地风电场测风塔数据的插补应用  

The Application of Meteorological Reanalysis Data for Wind Tower Data Interpolation at Complex Mountain Area Wind Farm

作　　者：韩二红 温新龙 王彬滨 杨富程[1] 沈宏涛[1] 祝铭宏 

机构地区：[1]中国电力建设集团四川电力设计咨询有限责任公司,成都610041 [2]江西省气象科学研究所,南昌330096

出　　处：《江西科学》2017年第2期200-205,234,共7页Jiangxi Science

基　　金：四川电力设计咨询有限责任公司科技项目(kj2015-f-7;kj2016-f-6);江西省重点科技成果转移转化计划项目"江西省风能资源详查及其成果应用"(20161BBI90040)

摘　　要：在四川盆周山地,风电场测风仪器因覆冰天气等因素导致失灵的现象时有发生。针对测风塔测风数据因覆冰天气等出现缺测的现象,使用高分辨率的气象再分析资料,对比了等权平均法、加权平均法、最优临近格点法、双线性插值法、反距离权重法以及神经网络(ANN)插值法对测风塔冬季结冰异常数据进行插补的效果。结果表明:神经网络(ANN)插值法的插补效果最优,双线性插值法与最优临近格点法效果差别较小,加权平均法则要优于等权平均法;通过对插补前后平均风速和威布尔分布的检验,发现神经网络(ANN)插值法具有较好的稳定性。In view of the wind tower data deficiency phenomenon occurring in winter in Sichuan province, this paper used high resolution meteorological reanalysis data and compared the effect of the abnormal data interpolation by the weight average method, the weighted average method, the optimal neighboring lattice point method, bilinear interpolation method, inverse distance weighting method and neural network （ANN） interpolation method for wind tower. This paper found that neural network （ANN） interpolation method is the optimal interpolation method. Bilinear interpolation method and the adjacent lattice point method have the same effect. The weighted average rule is superior to the weight average method. Based on the average wind speed and weibull distribution before and after the inspection, the neural network （ANN） interpolation method has good stability.

关 键 词：气象再分析资料 复杂山地风电场 测风数据 插补 

分 类 号：P49[天文地球—大气科学及气象学]