摘要
针对大部分区域空间降水数据无法直接获取以及传统降水插值方法的局限性,构建基于支持向量机的降水插值模型,并引入遗传算法优化模型中的惩罚参数和核函数参数,以江西省2015年1月和7月20个站台的降水数据作为数据源,通过不断更新迭代输入样本来改进模型,实现在少量数据的情况下分析模型的适应性以及插值精度。结果表明:基于遗传算法的支持向量机模型不仅在降水量较少和较多的月份均能进行降水插值,而且与传统方法相比精度有明显的提高。
Aiming at most of the regional precipitation data can not be obtained directly,and the limitation of the traditional interpolation method,this paper constructs a rainfall interpolation model based on support vector machine,and uses genetic algorithm to optimize the penalty parameter and kernel function parameter.By using the precipitation data of 20 stations in Jiangxi in January and July 2015,updating input data by constantly iterating,it can analyze the adaptability of the model and the interpolation accuracy in the case of a small amount of data.The results show that the support vector machine model based on genetic algorithm can not only perform the interpolate precipitation in less precipitation month and more precipitation month,but also improve the accuracy a lot compared with traditional methods.
作者
冯腾飞
钟钰
于良
Feng Tengfei;Zhong Yu;Yu Liang(School of Architectural and Surveying&Mapping Engineering,Jiangxi University of Science and Technology,Ganzhou Jiangxi 341000,China)
出处
《信息与电脑》
2017年第16期84-87,共4页
Information & Computer
关键词
降水插值
支持向量机
遗传算法
precipitation interpolation
support vector machine
genetic algorithm