基于粗糙集理论优化神经网络方法的轴承故障诊断

Bearing Breakdown Diagnosis of Rough Collection Theory Optimization Neural Network Method

用于轴承故障诊断方法有多种,如基频检测法、轴心轨迹图和响应信号的功率谱分析法等,但往往难以实现实时监测和诊断。人工神经网络技术由于其具有极强的非线性映射,能在故障诊断中得到广泛应用,但存在联想能力有限,超过界线易以错误的方式联想,决策系统就会产生误判或漏判的现象。提出了使用粗糙集理论优化BP神经网络模型方法,并将优化后的网络模型应用于滚动轴承的故障诊断中。

There are many kinds of the methods to diagnose the bearing breakdown,such as the method of base frequency test,the axle center trajectory diagram and the response signal power spectrum analytic method and so on,but it is often difficult to realize the real-time monitor and the diagnosis.Owing to the greatly strengthened non-linear insinuation,the artificial neural network technology can obtain the widespread application in the breakdown diagnosis,but the ability of association is limited.When it surpasses the demarcation line,it usually associates in the wrong way and the decision system will have the phenomena of misjudging or without judging.This article proposes a usage of rough collection theory by optimizing the BP nerve network model method,and will apply the optimized network model in the rolling bearing breakdown diagnosis.