【摘要】 针对旋转机械故障诊断中在同一征兆域中很难区分多种故障的实际情况 ,研究利用其他征兆域的诊断信息 ,进行全局信息融合 ,从而获得更为准确的故障定位 .提出一种基于神经网络和 Dempster-Shafer证据理论的多参数综合诊断系统模型 ;在该模型中采用证据理论的组合规则进行局部和全局信息融合 ,得出较好的诊断结果更多还原

【Abstract】 A model of multi parameter comprehensive diagnosis system based on neural networks and Dempster Shafer evidential reasoning was proposed with respect to the facts that it was difficult to distinguish different faults from one symptom domain during the fault diagnosis process of rotating machine. This paper presents the basic concepts and the evidence combining rules of Dempster Shafer evidential reasoning and induces the data fusion idea to the comprehensive fault diagnosis of rotating machine. More correct diagnosis results can be obtained by fusing part or full range information adopting the combining rules of evidence theory. At last the test by the actual example shows that the D S evidential reasoning method can be used in comprehensive fault diagnosis of complex system, and can make the judge more correctly.更多还原