基于深知识的故障应急系统模型

Deep Knowledge Based Fault Emergency System Model

系统故障的突发事件要求快速响应并处理,其特点是现象明显、故障原因复杂、波及的系统多、涉及的维护部门广,单凭经验很可能顾此失彼,而查找问题原因和确定相关技术人员常耗费大量时间。该文提出一种基于深知识的诊断模型,结合统计数据、专家经验利用模糊数学的方法,利用不确定性推理技术对不完备的知识进行不确定推理,在问题尚不明确的时候快速缩小范围,按概率推导出故障对象,最终得出最可能的故障原因、解决方案或给出最合适的维护人员,大大缩短了应急响应的时间。

Incidents of system failures require rapid response and disposal. Obvious phen-- omena and complex causes are characteristic of incidents which may affect many systems or in- volve different departments to deal with them. So depending on experience only probably takes one into consideration to the neglect of the other. At the same time, it may take a lot of time to search for causes of problems and to determine relevant technical staff. This paper presents a diagnosis model based on deep Knowledge. The model uses uncertainty reasoning technology to anal- ysis the incomplete knowledge, which makes problems decrease their range as soon as they are not dear. Fault targets as well as the most possible causes or solutions or the most appropriate ma- intenance personnel are deduced using probability, which significantly shortens the emergency response time.