基于神经网络的机械振动信号智能识别技术  

Intelligent Recognition Technology for Mechanical Vibration Signals Based on Neural Networks

作　　者：张玄 陈跃 王建国 梁樱紫 ZHANG Xuan;CHEN Yue;WANG Jianguo;LIANG Yingzi(Shandong Institute of Industrial Technology,Jinan 250000,China;Shandong Industrial Technology Research Institute Bozheng Innovation Consulting Co.,Ltd.,Jinan 250000,China;Yankuang Energy Group Company Limited,Jinning 273500,China;Jinan Shengquan Group Share Holding Co.,Ltd.,Jinan 250000,China)

机构地区：[1]山东产业技术研究院,山东济南250000 [2]山东产研博正创新咨询有限公司,山东济南250000 [3]兖矿集团,山东济宁273500 [4]济南圣泉集团股份有限公司,山东济南250000

出　　处：《智能物联技术》2024年第4期56-60,共5页Technology of Io T& AI

摘　　要：研究一种基于小波变换和神经网络的机械振动信号智能识别方法。构建基于神经网络的机械振动信号智能识别系统,利用小波变换对机械振动信号进行多尺度分解来提取信号的时频域特征向量,将其作为长短时记忆(Long Short-Term Memory,LSTM)模型的输入。利用凯斯西储大学轴承数据中心提供的数据集对LSTM模型进行训练与测试。结果表明,所提方法在故障类型识别中具有出色的表现,对正常状态和各类故障的分类准确率均高于95%,验证了基于小波变换和LSTM的故障诊断方法在实际应用中的有效性和可靠性。In this paper,an intelligent recognition method of mechanical vibration signal based on wavelet transform and neural network is studied.An intelligent recognition system of mechanical vibration signals based on neural network is constructed.The multi-scale decomposition of mechanical vibration signals is carried out by wavelet transform to extract the time-frequency domain feature vectors of the signals,which are used as the input of Long Short-Term Memory(LSTM)model.The LSTM model was trained and tested using the data set provided by the bearing data center of Case Western Reserve University.The results show that the proposed method has excellent performance in fault type identification,and the classification accuracy of normal state and all kinds of faults is higher than 95%,which verifies the effectiveness and reliability of the fault diagnosis method based on wavelet transform and LSTM in practical application.

关 键 词：振动信号 机械故障 小波变换 长短时记忆 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]