交叉熵神经网络及其在闽北大雨以上降水预报中的应用
作者:
基金项目:

福建省自然科学基金(2008J0241)


Application of BP neural network using cross-entropy to 96 hours forecast of heavy precipitation event in northern Fujian province
Author:
  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • | | |
  • 文章评论
    摘要:

    基于误差平方和最小化准则的BP神经网络(ANN-MSE)并不适合解决小概率天气事件的预报问题,引进一种改进的以交叉熵函数为目标函数的神经网络方法(ANN-CE),该法是一个三层反向传播神经网络,其输出层只用一个节点.利用2003-2008年的ECMWF预报场资料,把该法用于福建省南平市4-6月部分大雨或以上降水96h预报中,分别用原始因子和PCA降维后的主因子建立了ANN-CE预报模型和ANN-MSE预报模型,用这些模型对2009-2010年独立样本进行了试报.测试结果显示主因子预报模型TS评分比原始因子预报模型高且漏报次数少,其中,主因子ANN-CE预报模型的TS评分和漏报率分别是0.51和0.17,其性能是所有模型中最好且最为稳定的,是一种适合于小概率事件预报的方法.

    Abstract:

    As a neural network based on MSSE,ANN-MSE is not an appropriate solution to the problem of predicting rare weather event.In this paper,an improved neural network method,ANN-CE is presented,which is a three layered back-propagation neural network with one output unit.The error function of ANN-CE is a cross entropy function.Then utilizing ECMWF forecast fields data,this method was applied to 96 hours forecast of heavy precipitation event in northern Fijian province.The ANN-CE model and the ANN-MSE model based on original factor and principle component after PCA reducing dimensions were respectively built.These models were applied to independent samples in 2009-2010,and the test results are as following:TS grade for model based on principal component is higher than that of model based on original factors;miss-rate for the ANN-CE model is lower than that of the ANN-MSE model.All in all,ANN-CE model based on principal component has best performance and stability,whose TS grade and miss-rate was respectively 0.51 and 0.17,so it was suited for forecasting rare event.

    参考文献
    [1] 陆汉城.中尺度天气原理和预报[M].北京:气象出版社,2002:280-283LU Hancheng.Principle and forecast of medium rangeweather[M].Beijing:China Meteorological Press,2002:280-283
    [2] 施能.气象科研与预报中的多元分析方法[M].2版.北京:气象出版社,2002:1-5,72-82SHI Neng.Multivariate analysis for scientific research and forecast of meteorology[M].2nd Ed.Beijing:ChinaMeteorological Press,2002:1-5,72-82
    [3] 林健玲,金龙,林开平.神经网络方法在广西日降水预报中的应用[J].南京气象学院学报,2006,29(2):215-219LIN Jianling,JIN Long,LIN Kaipping.Application of ar-tificial neural network to daily regional average rainfallforecast in Guangxi[J].Journal of Nanjing Institute ofMeteorology,2006,29(2):215-219
    [4] 苗春生,段婧,徐春芳.人工神经网络方法在短期天气预报中的应用[J].江南大学学报:自然科学版,2007,6(6):648-653MIAO Chunsheng,DUAN Jing,XU Chunfang.Application of artificial neural network to forecast short-term pre-cipitation in Nanjing[J].Journal of Jiangnan University:Natural Science Edition,2007,6(6):648-653
    [5] 曹晓钟,闵晶晶,刘还珠,等.分类与集成方法在降雨预报中的应用[J].气象,2008,34(10):3-10CAO Xiaozhong,MIN Jingjing,LIU Huanzhu,et al.Ap-plication of classification and integration to rainfall forecast[J].Meteorological Monthly,2008,34(10):3-10
    [6] Ham F M,Kostantic I.叶世伟,王海娟译.神经计算原理[M].北京:机械工业出版社,2007:77-85Ham F M,Kostantic I.Principles of neurocomputing forscience&engineering[M].Columbus,OH:McGraw-Hill,2000
    [7] 金龙.神经网络气象预报建模理论方法与应用[M] 北京:气象出版社,2004:189-201JIN Long.Neural network modeling method&applicatiofor meteorological forecast[M].Beijing:China Meteorological Press,2004:189-201
    [8] 阎平凡,张长水.人工神经网络与模拟进化计算[M] 2版.北京:清华大学出版社,2005:437-447YAN Pingfan,ZHANG Changshui.Artificial neural networks and evolutionary computing[M].2nd Ed.BeijingTsinghua University Press,2005:437-447
    [9] Mitchell T M.曾华军,张银奎译.机器学习[M].北京机械工业出版社,2008:79-92,121-125Mitchell T M.Machine learning[M].Columbus,OHMcGraw-Hill,1997
    [10] 徐东,王翼飞.ANN-CE:一种预测DNA结合位点的改进神经网络方法[J].应用科学学报,2005,23(2):XU Dong,WANG Yifei.ANNCE:An improved neuranetwork method for predicting DNA binding sites[J] Journal Applied Sciences,2005,23(2):
    [11] Nasr G E,Badr E A,Goun C.Cross entropy function ineural networks:Forecasting gasoline demand[EB/OL]
    [2010-08-28].FLAIRS-02 Proceedings.http:∥wwwaaai.org/Papers/FLAIRS/2002/FLAIRS02-075.pdf
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

吴木贵,江彩英,张信华,赖荣钦.交叉熵神经网络及其在闽北大雨以上降水预报中的应用[J].南京信息工程大学学报(自然科学版),2012,4(3):220-225
WU Mugui, JIANG Caiying, ZHANG Xinhua, LAI Rongqin. Application of BP neural network using cross-entropy to 96 hours forecast of heavy precipitation event in northern Fujian province[J]. Journal of Nanjing University of Information Science & Technology, 2012,4(3):220-225

复制
分享
文章指标
  • 点击次数:1229
  • 下载次数: 2923
  • HTML阅读次数: 0
  • 引用次数: 0
历史
  • 收稿日期:2010-10-08

地址:江苏省南京市宁六路219号    邮编:210044

联系电话:025-58731025    E-mail:nxdxb@nuist.edu.cn

南京信息工程大学学报 ® 2025 版权所有  技术支持:北京勤云科技发展有限公司