基于LabVIEW和BP神经网络的旋转机械故障诊断研究

Research on fault diagnosis of rotating machinery based on LabVIEW and BP neural network

旋转机械产生的振动信号是复杂的非平稳信号,使用单一的BP神经网络,难以快速分析出故障的状态。该文研究了基于LabVIEW和BP神经网络的旋转机械故障诊断。利用LabVIEW软件作为测试平台,对旋转机械的振动信号进行分析和提取,构造出特征向量,以此作为BP神经网络的故障样本进行训练。实验结果表明,BP神经网络能够快速收敛到目标精度,为旋转机械故障状态诊断积累了应用经验。

The vibration signal produced by rotating machinery is a complex non-stationary signal.Using a single BP neural network,it is difficult to analyze the fault state quickly.This paper studies the fault diagnosis of rotating machinery based on LabVIEW and BP neural network.Using LabVIEW software as the test platform,the vibration signal of rotating machinery is analyzed and extracted,and the eigenvector is constructed,which is used as the fault sample of BP neural network for training.Finally,the experimental results show that the BP neural network can quickly converge to the target accuracy,which accumulates the application experience for the fault diagnosis of rotating machinery.