位置:成果数据库 > 期刊 > 期刊详情页
燃料电池发动机故障诊断专家系统设计与研究
  • 期刊名称:武汉理工大学学报(交通科学与工程版)
  • 时间:0
  • 页码:280-284
  • 语言:中文
  • 分类:U468.3[机械工程—车辆工程;交通运输工程—载运工具运用工程;交通运输工程—道路与铁道工程]
  • 作者机构:[1]武汉理工大学汽车工程学院,武汉430070, [2]武汉理工大学自动化学院,武汉430070
  • 相关基金:国家自然科学基金项目(批准号:60705032/F030605 61004018/F030102); 国家高技术研究发展计划资助项目(批准号:2008AA05Z105); 中央高校基本科研业务费专项资金(校内批准号:2010-Ⅱ-012)资助
  • 相关项目:燃料电池电动汽车机电混联式柔性制动关键问题研究
中文摘要:

针对自主研发的60kW车用燃料电池发动机的可靠性与安全性要求,结合实际研发经验和燃料电池的工作机理,设计了其基于故障树的故障诊断专家系统.根据其常见故障模式,建立了系统的故障树模型,用字段和代码表示各节点的树状层次关系从而构建知识库,通过模糊产生式规则对知识进行表达,采用正向推理方式实现系统的推理,并基于VC++语言开发出了燃料电池发动机的故障诊断专家系统.该系统人机界面友好、知识库维护方便,具有一定的自学习功能.最后,以氢气供给子系统失效为诊断实例,阐述了系统的故障诊断过程并验证了其有效性和实用性.

英文摘要:

According to the reliability and security performance requirements of 60kW automotive fuel cell engine(FCE) designed by our group,by combining our practical development experience with the operation mechanism of fuel cells,its fault diagnosis expert system based on fualt tree was designed in this paper.According to its common fault modes,the fault tree model of FCE system was set up,its hiberarchy relationship among each nodes was denoted with words field plus codes to establish the knowledge base,and the knowledge was expressed with fuzzy production rule,then the fault diagnosis expert system of FCE was realized with VC++ language.The man-machine interacted interface of the expert system is friendly,its knowledge base is maintainable,and it also has good self-learing function.At last,taking the hydrogen supply sub-system malfunction for example,its fault diagnosis process was addressed,which proves that the expert system developed is effective and practical.

同期刊论文项目
同项目期刊论文