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

芽胞气溶胶表面滞留抗力预测模型研究
引用本文:黄国荣,张林,王庆华,向颖,许斌,熊鸿燕.芽胞气溶胶表面滞留抗力预测模型研究[J].微生物学杂志,2010,30(5):19-24.
作者姓名:黄国荣  张林  王庆华  向颖  许斌  熊鸿燕
作者单位:第三军医大学,军队流行病学教研室,重庆,400038
基金项目:重庆市科委攻关课题,第三军医大学军事预研基金课题 
摘    要:针对生物威胁的现场处置工作,建立气溶胶芽胞表面滞留抗力的智能预测模型,以准确预测环境表面芽胞污染状况,为大规模的现场洗消任务提供重要依据,有利于实现及时反应、恰当反应和准确防护的目标。以枯草杆菌芽胞为试验菌,在气溶胶实验室进行芽胞的环境因素暴露及活力测定,以模拟环境中芽胞抗力变化规律数据为依据,采用Matlab6.1软件包中的神经网络工具箱进行抗力预测模型研究。根据研究目的、模拟环境条件和数据训练的平滑曲线等特征,设定了5个输入神经元,8个隐层节点和1个输出神经元。‘tansig’、‘purelin’为传递函数,trainlm为训练函数,网络迭代100次。模型回顾预测效率达到100%,前瞻预测效率达到91%。以实验室数据为依据,利用Matlab平台中的BP神经网络建立的芽胞气溶胶表面滞留抗力预测模型能利用环境因素信息有效预测芽胞抗力。

关 键 词:芽胞  抗力  预测模型

The Study on Intellectual Forecasting Model for Resistance of Spore Aerosol Staying on Object Surface
HUANG Guo-rong,ZHANG Lin,WANG Qing-hu,XIANG Ying,XU Bin,XIONG Hong-yan.The Study on Intellectual Forecasting Model for Resistance of Spore Aerosol Staying on Object Surface[J].Journal of Microbiology,2010,30(5):19-24.
Authors:HUANG Guo-rong  ZHANG Lin  WANG Qing-hu  XIANG Ying  XU Bin  XIONG Hong-yan
Institution:(Chongqing Third Military Medical University,Chongqing,400038)
Abstract:Directing against the work to handle a bio-threat on the site,an intellectual forecasting model regarding resistance of spore aerosol staying on object surface was established in order to accurately predict the contamination conditions of spore on environmental surface and provide foundation for washing and sterilizing the site and conducive for bringing about in-time reaction,proper reaction and accurate protection.Bacillus subtilis was used as a test bacterium in the lab to carry out determination of the environmental factors and the spore activities to provide a foundation for simulating the law of spore resistance changes in the environment,an artificial neural network(ANN) in Matlab 6.1 software was adopted to carry out the model study on resistance prediction.The results showed that according to the goal of the study,simulative environmental conditions and smooth curve of data training and other features,five input neurons,eight hidden layer nodes and one output neurons were set up.Took 'tansig','purelin' as transfer functions,"trainlm" as training function,and network iteration 100 times.The forecasting efficiency of model retrospective prediction reached as high as 100%,and the efficiency of prospective prediction was reached as high as 91%.Therefore,based on lab experimental data using prediction model of spore aerosol on surface staying resistance established with BP ANN tools on platform of Matlab could effectively forecast the spore resistance using environmental factor information.
Keywords:spore  resistance force  forecasting model
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《微生物学杂志》浏览原始摘要信息
点击此处可从《微生物学杂志》下载免费的PDF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