论文部分内容阅读
本文提出一种神经网络与模糊逻辑相结合的故障诊断系统,该系统包括2个方面:模糊推理模块和规则学习模块。模糊推理规则记忆在网络的记忆层中,记忆节点的激活水平则反映了输入矢量与已记忆规则的匹配程度;规则学习模块通过自组织聚类过程自动生成规则。作为该诊断系统的一个应用实例,模拟了旋转主轴的故障诊断试验。
This paper presents a fault diagnosis system based on neural network and fuzzy logic. The system includes two aspects: fuzzy reasoning module and rule learning module. The fuzzy inference rules are stored in the memory layer of the network, and the activation level of the memory nodes reflects the match between the input vector and the already-stored rules. The rule learning module automatically generates rules through a self-organizing clustering process. As an application example of this diagnostic system, the fault diagnosis test of rotating main shaft is simulated.