柴油机磨损状态监测及故障诊断专家系统领域知识的获取

Acquisition of the Expert System Field Knowledge for Engine Wear Condition Monitoring and Fault Diagnosis of Dielsel

目前，摩擦学系统分析与摩擦学设计是摩擦学领域中很受重视的研究方向之一。继柴油机磨损状态监测及故障诊断专家系统ＦＤＥＸＳＹＳ（Ｖ１．０）知识库建立的研究之后，在摩擦学系统的系统工程思想指导下，重点研究了ＦＤＥＸＳＹＳ专家系统领域知识的获取。以铁谱技术、光谱技术和润滑油性能指标测定技术作为磨损状态监测及故障诊断的工具，研究了由此输出的信息而获取领域知识。由于摩擦学知识结构的特殊性，对领域知识的获取综合了多种来源的知识，同时还考虑了知识来源的可靠性对其进行了定量分析。在此基础上，明确提出了柴油机磨损状态监测及故障诊断知识的来源与分类，增强了知识的模块性，可在利用计算机实现摩擦学过程的在线监测与诊断方面推广应用。

At present,the triblogical systematic analysis and tribological design is one of the research direction to be attached importance in the tribological field. After the knowledge bank ofadiesel engine wear condition monitoring and fault diagnosis expert system FDEXSYS(V1.0)hasbeen established, the focus of investigation should be turned to the task of field knowledgeacquisition。In this paper,information concerning with the wear condition of the engine comes from ferrograph,spectrograph and other physical/ chemical measurements.How to transform such information into field knowledge is discussed.Because of the speciality of the tribology knowledge structure, synthesis the knowledge from different sources is necessary。 The reliability of every knowledge piece must be considered,therefore some quantitative analysis has been carried out.On the base of analysis,the sources of diesel engine wear condition monitoring and fault diagnosis knowledge and their categories are clearly suggested and the modularization of knowledge structure has been improved.The results can be applied to the online monitoring and diagnosis of tribo process with computers。