基于PCA-RBF的风电场短期风速订正方法研究  

Correction Method of Short-Term Wind Speed in Wind Farm Research Based on PCA and RBF Neural Network

作　　者：邓华[1,2,3] 张颖超 顾荣[1,2] 黄飞[1,2] 支兴亮 

机构地区：[1]南京信息工程大学气象灾害预报预警与评估协同创新中心,南京210044 [2]江苏省大数据分析技术重点实验室,南京210044 [3]气象灾害教育部重点实验室,南京210044

出　　处：《气象科技》2018年第1期10-15,共6页Meteorological Science and Technology

基　　金：国家自然科学基金项目(41675156);国家公益性行业(气象)科研专项(GYHY20110604);江苏省六大人才高峰项目(WLW-021)资助

摘　　要：风电功率预测中最重要的因子是风速,准确的风速预测是风电功率预测的前提和基础。为了提高短期风速预测的准确性,本研究采用WRF模式,对我国上海崇明吕四风电场的风速进行预报。在此基础上,利用PCA-RBF算法结合WRF模式预报风向、气温、气压等气象要素对预报风速进一步订正。实验结果表明,利用PCA-RBF算法对WRF模式预报风速进行订正后,预报风速的误差进一步减小,相对均方根误差降低20%~30%,相对平均绝对误差降低15%~20%。与其他智能算法(BP算法、LSSVM算法)对比分析后得出,PCA-RBF算法对WRF模式预报风速具有较好的订正效果,能够有效提高风速预报准确率。Wind speed is the most important input factor of wind power forecasting,and the accurate wind speed forecasting is the premise and foundation of wind power prediction.In order to improve the accuracy of short-term wind speed forecasting,the WRF model is used to predict the wind speed of a wind farm along the east coasts of China.Besides,the WRF model forecasted wind direction,air temperature,barometric pressure and other meteorological factors are combined by the PCA-RBF algorithm to further correct the forecasting wind speed.The results show that,after the correction of the PCA-RBF algorithm for the wind speed forecasting of the WRF model,the error of wind speed forecasting becomes smaller,and the relative root mean square error is reduced by 20% to 30%,and the relative mean absolute error is decreased by 15% to 20%.The PCA-RBF algorithm is qualified with better correction for the wind speed of WRF model forecasting compared with other intelligent algorithms（BP algorithm,LSSVM algorithm）,and improves the accuracy of wind speed forecasting effectively.

关 键 词：WRF模式 PCA算法 RBF算法 风速订正 

分 类 号：P425[天文地球—大气科学及气象学]