【摘要】 为提高机械维修人员故障诊断效率,提出了可对维修过程实时指导的智能语音交互的维修型专家系统.根据诊断流程适合采用树状结构表示的特点,提出逐级诊断和跨越诊断两种模式,并设计了相应的专家系统推理机制.借助讯飞语音输入法方言识别功能,可实现灵活的语音交互.针对跨越诊断模式,提出了适合的文本比较算法——Needleman/Wunsch算法.研究了专家系统开发过程中的关键技术,提出了改进的CLIPS和VC++交互方法.以某推土机故障为案例,采用专家系统工具CLIPS和VC++结合开发推土机故障诊断专家系统.推土机语音故障诊断专家系统表明,可通过智能语音交互从而解放维修人员双手,对提高维修效率有一定帮助,且软件兼容性能良好,可运行于Windows平板电脑上,方便维修人员在狭窄空间携带.更多还原

【Abstract】 To improve the efficiency of maintenance staff in fault diagnosis,fault diagnosis expert system based on intelligent voice interaction for live guide was proposed.According to the features of procedure of diagnosis appropriate to be represented by tree structures,two modes of layer-by-layer diagnosis and skiplevel diagnosis were presented,and inference engine of expert system is designed in this paper.Using accent recognition function of iFlytek Input can realize complex speech interaction.According to skip-level diagnosis,the paper studied arithmetic of text comparison on all longest common string—Needleman/Wunsch.The key technologies of expert system was studied.The paper discussed the interoperation technology of VC++ and CLIPS,and come up with a method of improved interoperation technology.This paper develop fault diagnosis expert system of bulldozer by applying VC++ and CLIPS interaction method based on fault diagnosis of bulldozer.Bulldozer fault diagnosis expert system based on intelligent voice interaction indicated hands of maintenance staff freed and the efficiency of fault diagnosis improved.The results showed that the software has good compatibility for running in PAD based on Windows operating system and PAD is easy to carry in limited space.更多还原