基于不确定推理的数控机床故障诊断的研究  

Study on CNC fault diagnosis based on uncertainty reasoning algorithm

作　　者：王家海[1] 刘晨阳[1] 

机构地区：[1]同济大学机械与能源工程学院,上海201804

出　　处：《制造技术与机床》2014年第1期48-52,共5页Manufacturing Technology & Machine Tool

基　　金：国家科技部科技重大专项项目:高速复合数控机床及关键技术创新能力平台(2011ZX04016-21)

摘　　要：数控机床一旦发生故障,因其结构复杂性往往花费大量时间寻找根本故障原因。为实现数控机床故障原因的快速准确定位,针对当前数控机床故障诊断技术在实际应用中遇到的问题,提出了一种基于报警号码优先级的不确定推理算法。该算法结果能够随着报警号码优先级的不同而产生变化,并重点分析了不同计算顺序对改进算法可信度计算结果的影响,提出了一种较为合理的计算顺序。最后,以实际数控系统故障为实例,说明了改进算法在数控机床故障实例诊断中的应用,证实了可行性。Once CNC machine tools break down,it will spend lots of time to find out root cause usually because of its complex structure. Due to practical problems of fault diagnosis for CNC machine tools,an uncertainty reasoning algorithm using alarm priority is presented in order to realize the rapid location of CNC fault cause. The result of this algorithm can be changed with the difference of alarm priority,then the influence of different computation sequence on computed result is analyzed with emphasis and one reasonable sequence is presented. Finally take actual CNC fault for instance,the application of this algorithm in CNC machine tools fault diagnosis is illustrated and its validity is proved.

关 键 词：不确定推理 数控机床 故障诊断 

分 类 号：TH165[机械工程—机械制造及自动化] TP311[自动化与计算机技术—计算机软件与理论]