基于神经网络的滚动轴承故障诊断方法的研究

神经网络是一种具有非线性映射能力强以及自学习、自组织、自适应等优点的智能方法,非常适合于滚动轴承的故障诊断。针对滚动轴承是机械设备重要的易损零件之一,大约有30%的故障是由轴承损坏引起的,提出了基于神经网络的滚动轴承故障诊断方法。以滚动轴承小波分解后的能量信息作为特征,通过神经网络作为分类器对滚动轴承故障进行识别、诊断。实验表明,该方法对于滚动轴承的故障诊断具有良好的效果和应用价值,并可方便地推广到其他类似的诊断领域。

Fault diagnosis of rolling bearing based on neural network

Neural network is a non-linear intelligent method has the advantage of a mapping capability and self-learning,self-organizing,adaptive,etc,and which is ideal for rolling bearing fault diagnosis.We proposed a method for fault testing on rolling bearing based on neural network in connection with rolling bearing is important to machinery and equipment of fragile parts and approximately 30% of which failures are caused by bearing damage.The method use the energy information after rolling wavelet decomposition as a feature and the neural network as classifier to identify and diagnosis the rolling bearing fault.Experiments show that the method is the fault diagnosis for rolling bearing with good results and value,and can be easily extended to other similar diagnostics.