VC调用Matlab的出口食品风险分析的神经网络方法  

Neural Network Method in Risk Analysis of Exported Seafood Based on Mixed Programming

作　　者：黄金明[1] 宋明慧[1] 武玉强[1] 王文增 

机构地区：[1]曲阜师范大学电气信息与自动化学院 [2]日照出入境检验检疫局

出　　处：《电子技术（上海）》2011年第9期1-3,共3页Electronic Technology

基　　金：曲阜师范大学校级项目资助(编号:XJ201022);山东省高等学校科技计划资助项目(编号:J11LG08);山东出入境检验检疫局科技攻关资助项目(编号:SK200834)

摘　　要：出口海产食品的传统定性检验结论不能体现产品质量差异对风险的影响。文章将神经网络引入海产食品风险分析过程中,利用VC调用Matlab的DLL方法,对海产食品出口的风险度进行了定量评估。比较了VC和Matlab各自的优缺点,详细阐述了神经网络的建模方法以及二者混合编程的步骤。测试结果表明该方法具有较高的精确度和可行性,为海产食品的出口风险评估提供了一种新的解决方案。At present the inspection conclusions of exported seafood with traditional qualitative inspection can not reflect the influence of products’ quality difference on the risk.This paper introduces BP-neural network into the risk analysis process of seafood,and uses DLL method which calls Matlab using VC to make a quantitative assessment on the risk of exported seafood.The advantages and disadvantages of VC and Matlab are compared,and the modeling method of neutral network and the procedure of the mixed programming of them are described in detail.The test results show that the method has higher accuracy and feasibility,and provides a new solution for the exporting risk assessment of seafood.

关 键 词：BP神经网络 风险分析 MATLAB 混合编程 非线性 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]