轴承故障的智能诊断方法

Intelligent Diagnosis Method of Bearing Fault

总结了轴承故障诊断的模糊规则 ,并把这些规则映射到模糊神经网络 ,提高了诊断的智能性。提出了部分层学习算法 ,并推导出隶属度函数的参数学习算法 ,改善了诊断规则和学习性能。所研制的诊断系统在对轴承故障的诊断中取得了满意的结果。

Fuzzy rules of bearing fault diagnosis are summarized and maped into fuzzy neural network, which improves intelligence of bearing diagnosis. Part layer learning algorithm is presented. Meanwhile, parameter learning algorithm of the membership function is developed. Both of them improve diagnostic rules as well as learning properties. As to bearing fault, satisfied diagnosis results are obtained by means of the developed diagnostic system.