首页 | 本学科首页   官方微博 | 高级检索  
     检索      

用支持向量机方法预报舟山沿海突发性冷空气大风
引用本文:陈淑琴,张蔺廉,黄辉.用支持向量机方法预报舟山沿海突发性冷空气大风[J].海洋预报,2007,24(1):51-55.
作者姓名:陈淑琴  张蔺廉  黄辉
作者单位:浙江省舟山市气象局,舟山,316021
摘    要:本文用舟山市1994~2003年的测风站资料统计了突发性冷空气大风的发生规律,是以中路冷空气为主,主要发生在11月到3月。通过t统计分析,认为与冷空气大风关系密切的物理量主要有各层风速、低层的温度梯度、地面的气压梯度等,因此主要从这些物理量中选取预报因子。用SVM方法建立预报模型,结果证明SVM方法确实适合小样本建模,有较好的推广能力。

关 键 词:支持向量机  突发性  冷空气大风
文章编号:1003-0239(2007)1-0051-05
修稿时间:2006-02-23

Forecasting sudden cold air gales by a method of support vector machine
CHEN Shu-qin,ZHANG Lin-lian,HUANG Hui.Forecasting sudden cold air gales by a method of support vector machine[J].Marine Forecasts,2007,24(1):51-55.
Authors:CHEN Shu-qin  ZHANG Lin-lian  HUANG Hui
Institution:Zhoushan Meteorological Bureau, Zhoushan 316021 China
Abstract:The winds data of Zhoushan from 1994 to 2003 are analyzed to find the rules of occurrence of sudden gales caused by cold airs,the result show that most of which are caused by cold airs from the middle area and are take place from November to February.According to the analyses of T statistics,it is these things that have close relation to cold air wind such as winds at all levels,temperature gradient at low levels,pressure gradient on surface,and so on.So forecast factors are selected out of these things.At last,a forecast model is established by a method of support vector machine.The result proves that this method is good at establishing model by a few samples and has the good capability of spread.
Keywords:support vector machine  suddenness  strong gales caused by cold air
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号