摘要
利用LabVIEW图形化编程语言开发了信号分析与处理、信号特征提取和故障诊断三大模块。信号特征提取由小波包分解来实现,故障诊断通过神经网络完成,小波包分解提取的齿轮振动信号各频段能量特征值作为神经网络的输入向量。以模拟故障实验台获取的齿轮典型故障振动信号训练神经网络,利用训练好的神经网络对齿轮进行故障诊断,实验结果表明:所开发的齿轮故障智能诊断系统能有效识别齿轮故障,较好地将虚拟技术应用于故障诊断领域。
The three modules like the signal analysis and processing,the signal feature extraction and the fault diagnosis were developed with LabVIEW's graphical programming language,in which,the wavelet packet decomposition can implement signal feature extraction,and the neural network completes fault diagnosis,as well as the feature of every frequency segment energy of the gear vibration signal picked up by the wavelet packet decomposition can be taken as the input vector of neural network.The neural network to be trained with gear's typical fault vibration signal acquired through imitation fault testing platform was employed to diagnose the gear fault.The experimental result proves the effectiveness of the intelligent diagnosis system for gear faults and the applicability of virtual technology in fault diagnosis.
出处
《化工自动化及仪表》
CAS
2013年第6期762-765,共4页
Control and Instruments in Chemical Industry
基金
教育部重点实验室资助项目--载运工具与装备(11TD07)