基于模糊数学的决策树技术在信号知识库更新中的应用  

The Application of Decision Tree Technology in Signal Knowledge Based on Fuzzy Mathematics

作　　者：胡岳鹏[1] 李淑芳[1] 

机构地区：[1]兰州交通大学自动化与电气工程学院,甘肃兰州730070

出　　处：《科技创新与生产力》2011年第4期100-102,105,共4页Sci-tech Innovation and Productivity

摘　　要：针对当前铁路信号设备维修中维修信息的模糊性、信息获取的不完备性,使维修决策支持系统知识库的更新变得很困难这个问题,提出了通过对知识的模糊性表示,采用决策树分类技术对设备状态进行分类,形成具有信息确信度的子树,通过子树合成形成最终的决策树,最后从该决策树中提取模糊性产生式规则,从而对信号设备的维修决策支持系统知识库中的知识进行更新,以提高信号维修决策支持系统的决策分析能力。In this paper,aiming at problems of ambiguity and incompleteness theorem of servicing information in servicing railway signals equipment,the auhtor classified equipment status by using classification technique of decision tree in order to forming subtree that has certainty factor of information through knowledge ambiguity.Through forming ultimating decision,the author extracted fuzzy production rules to update knowledge base,which could improve analysis ability of servicing supporting system.

关 键 词：决策树 知识库 模糊产生式规则 信息增益 训练数据集 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]