首页 | 本学科首页   官方微博 | 高级检索  
     

基于径向基函数网络的溢油预测模型
引用本文:周慧, 陈澎. 基于径向基函数网络的溢油预测模型[J]. 环境工程学报, 2014, 8(11): 5062-5066.
作者姓名:周慧  陈澎
作者单位:1.大连东软信息学院计算机系, 大连 116023; 2.大连海事大学航海学院, 大连 116023
基金项目:国家自然科学基金资助项目(41171329)
摘    要:为了提高溢油预测的准确性,建立和优化溢油预测模型,提出了基于径向基函数网络模型的溢油预测方法,实现溢油预测功能.径向基函数网络模型解决了模拟预测过程中样本库巨大、函数模型收敛速度慢的问题.通过选择有效的输入参数和样本数据,建立局部逼近网络;通过径向基函数训练样本数据,利用输出值与实际值之间的误差作为约束条件调整权重因子、径向基中心和宽度,加快函数模型的收敛速度.该模型模拟了溢油的漂移、扩散过程,达到预测的目的.利用该模型,建立了溢油预测模块,并针对一次溢油事故进行预测模拟,验证了该模型的可行性,能够为应急决策提供一定的支持.

关 键 词:溢油预测   径向基函数网络   溢油预测模型   模型设计   应急决策
收稿时间:2014-03-13

Oil spill prediction model based on radial basis function network
Zhou Hui, Chen Peng. Oil spill prediction model based on radial basis function network[J]. Chinese Journal of Environmental Engineering, 2014, 8(11): 5062-5066.
Authors:Zhou Hui  Chen Peng
Affiliation:1.Department of Computer Science, Dalian Neusoft University of Information, Dalian 116023, China; 2.Navigation College, Dalian Maritime University, Dalian 116023, China
Abstract:We propose a new prediction method based on RBF network in order to improve the prediction accuracy. RBF network model is used to solve the problem of massive sample set and slow convergence rate in simulating the prediction process. We set up the local approximation network through selecting effective input parameters and sample data. The RBF network is used to train the data and the convergence rate is improved through adjusting the weighting parameters, center and width of the radius basis, according to the constraints built based on the difference between the output values and the actual values. And this model can realize oil spill prediction through simulating the drift and diffusion process. We build the prediction module using the proposed model and apply it to real-world experiment. The experimental results demonstrate that the proposed model can supply reliable decision in emergency.
Keywords:oil spill prediction  radial basis function network  the model of oil spill prediction  model design  emergency decision-makings
本文献已被 CNKI 等数据库收录!
点击此处可从《环境工程学报》浏览原始摘要信息
点击此处可从《环境工程学报》下载全文
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