摘要
提出了故障多征兆域一致性的诊断思想,并对该思想的实现途径进行了研究;提出了从群体专家一致性决策角度基于BP神经网络实现故障多征兆域一致性诊断的思路,且具体研究了基于Dem pster-Shafer证据理论和基于模糊积分理论实现一致性处理的两种有效策略;最后就所提出的方法和策略进行了示例研究。
Following the idea of “dispersion first, then concentration” adhering to the biological perception system, the idea of the multi symptom domain fault diagnosis is developed. From the view point of group decision, measures based on neural networks to realize this diagnosis idea are studied, and two particular multi symptom domains diagnosis strategies are proposed, which are based on Dempster Shafer evidence theory and Fuzzy integral theory respectly. Finally, a case study is given. The research results show that the diagnosis strategy proposed is more efficient than conventional methods.
出处
《振动工程学报》
EI
CSCD
1999年第4期447-453,共7页
Journal of Vibration Engineering
基金
国家自然科学基金!(59275240)
关键词
故障诊断
D-S
证据理论
模糊积分理论
多征兆域
fault diagnosis
neural network
evidence theory
fuzzy integrals theory
group decision
