基于支持向量机回归的应急物资需求预测
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
目前对应急物资需求预测仍以专家经验判断为主,尚未形成成熟的办法。针对应急物资预测方法缺乏科学依据难题,提出了一种新的科学预测方法 -两步法,即支持向量机回归算法和库存管理模型,建立了人员伤亡预测模型,对地震人员伤亡进行了预测,然后再结合库存管理模型对应急物资进行了估算。最后,运用提出的方法对"青海玉树地震"的人员伤亡进行了预测,证明了方法的有效性,并估算出了应急物资需求量,为灾区的应急物资供应提供科学依据。
A new scientific method was established based on two-steps method—the Support Machine Vector Regression and the Inventory Management Model,which can predict the earthquake casualties.Next,the emergency materials were estimated by the Inventory Management Model.At last,the casualties of"Qinghai Yushu earthquake" were predicted by this method.The method was proved effective.Next,the emergency materials of it were estimated,and this provided the scientific basis for the supply of emergencies.
引文
[1]郑宇.城市防震减灾能力评价指标与应急需求研究[D].南京工业大学,2003.
    [2]宋晓宇,等.基于改进GM(1,1)模型的应急物资需求量预测[J].沈阳建筑科技大学学报,2010,26(6):6-8.
    [3]郭金芳,等.大型地震应急物资需求预测方法研究[J].ValueEngineering,2011,22:2-3.
    [4]傅志研,等.灾害应急物资需求预测模型研究[J].物流科技,2009,(10):11-13.
    [5]VAPNIKV.The Nature of Statistical Learning Theory[M].NewYork:Spring Verlag,1995.
    [6]白鹏,张喜斌,张斌,李彦,谢文俊,刘君华.支持向量机理论及工程应用实例[M].西安:西安电子科技大学出版社.2008,53-5 7.
    [7]Jiuh-Bing Sheu.An emergency logistics distribution approach forquick response to urgent relief demand in disasters[J].Transporta-tion Research Part E,2007,43:687-709.

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心