基于模糊神经网络的自动生产线故障诊断方法研究  

Research on Auto-production Line Fault Diagnosis Based on FNN

作　　者：吴宗彦[1] 韩煜[1] 张建军[1] 张利[1] 

机构地区：[1]合肥工业大学,合肥230009

出　　处：《中国机械工程》2008年第9期1072-1075,共4页China Mechanical Engineering

摘　　要：针对自动生产线故障的特点,选择了利用模糊神经网络对其进行故障诊断的方法,并给出了进行故障诊断的总体方案。在此基础上,建立了基于模糊神经网络与信息融合技术的自动生产线故障诊断模型。为了验证该模型,结合某球轴承套圈磨削超精自动生产线,运用MATLAB神经网络工具箱对该生产线上数控内圆磨床砂轮部分的故障数据进行建模、仿真与测试,结果表明,该模型能快速、准确、有效地诊断出故障。According to the failure characteristics of auto--production line, this paper adopted FNN as fault diagnosis method and designed a overall scheme for it. A diagnosis model of auto--production line based on FNN and data fusion technology was established. In order to test this model, this paper combined with a automatic super finishing line for ball bearing ring grinding, used the MATLAB neural network tool box to simulate and test the model which was established to diagnose the grinding wheel failure of the NC internal grinder on the auto--production line. The results show that model can diagnose the faults rapidly, accurately and effectively.

关 键 词：模糊神经网络 自动生产线 故障诊断 信息融合 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]