模糊神经网络故障诊断专家系统

Faults Diagnosis System Based on Fuzzy Neural Network

模糊诊断和神经网络有机结合能够提高设备故障诊断技术水平，文章阐述了单症兆诊断模型及多症兆诊断模糊模型与神经网络的有机结合；论述以模糊诊断原理为根据发展的扩展神经网络子网和子节点的可塑性学习方法和以模糊模式识别理论指导学习样本的组织方法。简述基于上述模糊神经网络的诊断系统及诊断系统的生产现场实例验证。

Combing fuzzy theory with neural network may improve the technology of mechanical machinery fault diagnlsis. This paper demonstrates the combination of diagnlsis models of single symptom and multi-symptom with neural networks. According to fuzzy diagnosis model, this paper supposes a plastic learning method by adding subnetworks and subjoints. Fuzzy pattern recognition theory directs the samples orgination of neural learning. Based on fuzzy neural network model,a machinery fault diagnlsis system is built and tested in production plant.