神经网络在毒剂识别中的应用  

The Application of Neural Networks in Chemical Agent Identification

作　　者：王洛国[1] 

机构地区：[1]中国人民解放军防化指挥工程学院四系,北京102205

出　　处：《太原理工大学学报》2008年第6期603-605,642,共4页Journal of Taiyuan University of Technology

摘　　要：识别毒剂方法已有很多,但自动化、智能化的程度都很低。为快速、准确地识别毒剂,在分析神经网络毒剂识别基本方法的基础上,建立了带有偏差单元的递归神经网络毒剂识别模型,设计了神经网络毒剂识别的软件,实现了神经网络对毒剂的识别。用沙林模拟数据对该模型进行了分析和测试,得到了比较满意的结果。Many methods have been developed for chemical agent identification, but all with low automation and intelligentization. For the purpose of quick and accurate identification of chemical agents,a recurrent neural network model with deviation unit was proposed on the basis of analyzing the basic methods for neural network identification of chemical agents, and the corresponding software was designed to realize the identification of chemical agents by neural network. A satisfactory result was obtained by analyzing and testing the model with simulated GB data.

关 键 词：神经网络 特征提取 毒剂识别 

分 类 号：E929[军事—军事装备学]