摘要
本文提出了一种基于一维卷积神经网络对齿轮箱进行故障诊断的方法;建立了一维卷积神经网络结构模型;优化了网络参数;设计了基于工程数据源的实验方案;探究了一维卷积神经网络对齿轮不同故障的分类准确度.实验表明:在识别齿轮箱的故障模式的过程中,一维卷积神经网络能准确区分齿轮的故障与正常状态,较为准确地分类出单独故障,但对于复合故障的分类能力下降.
A method of gear box fault diagnosis based on one-dimensional convolution neural network was proposed.The model of one-dimensional convolution neural network was established,the network parameters were optimized.The experimental scheme based on engineering data source was designed,and the classification accuracy of one-dimensional convolution neural network for different gear faults was explored.In the process of identifying the fault modes of gearbox,one-dimensional convolution neural network can accurately distinguish the fault and normal state of gears,less accurately classify the individual fault,and the ability to classify complex fault is reduced.It provided a theoretical basis for gear fault diagnosis using one-dimensional convolutional neural networks.
引文
[1]程军圣,杨怡,杨宇.局部特征尺度分解方法及其在齿轮故障诊断中的应用[J].机械工程学报,2012,48(9):64-71.Cheng Junsheng,Yang Yi,Yang yu.Local characteristic-scale decomposition method and its application to gear fault diagnosis[J].Journal of Mechanical Engineering,2012,48(9):64-71.(in Chinese)
[2]刘秀丽,徐小力.基于深度信念网络的风电机组齿轮箱故障诊断方法[J].可再生能源,2017,35(12):1862-1868.Liu Xiuli,Xu Xiaoli.Fault diagnosis method of wind turbine gearbox based on deep belief network[J].Renewable Energy Resources,2017,35(12):1862-1868.(in Chinese)
[3]李东东,王浩,杨帆,等.基于一维卷积神经网络和Soft-Max分类器的风电机组行星齿轮箱故障检测[J].电机与控制应用,2018(6):80-88.Li Dongdong,Wang Hao,Yang Fan,et al.Fault detection of wind turbine planetary gear box using 1Dconvolution neural networks and soft-max classifier[J].Electric Machines&Control Application,2018(6):80-88.(in Chinese)
[4]He K,Sun J.Convolutional neural networks at constrained time cost[C].Computer Vision and Pattern Recognition.IEEE,2014:5353-5360.
[5]田娟,李英祥,李彤岩.激活函数在卷积神经网络中的对比研究[J].计算机系统应用,2018,27(7):43-49.Tian Juan,Li Yingxiang,Li Tongyan.Contrastive study of activation function in convolutional neural network[J].Computer Systems&Applications,2018,27(7):43-49.(in Chinese)
[6]陈旭,张军,陈文伟,等.卷积网络深度学习算法与实例[J].广东工业大学学报,2017(6):20-26.Chen Xu,Zhang Jun,Chen Wenwei,et al.Convolutional neural network nlgorithm and case[J].Journal of Guangdong University of Technology,2017(6):20-26.(in Chinese)
[7]Boureau Y L,Ponce J,Lecun Y.A Theoretical Analysis of Feature Pooling in Visual Recognition[C].International Conference on Machine Learning.DBLP,2010:111-118.
[8]Sainath T N,Mohamed A R,Kingsbury B,et al.Deep convolutional neural networks for LVCSR[C].IEEE International Conference on Acoustics,Speech and Signal Processing.IEEE,2013:8614-8618.
[9]李彦冬,郝宗波,雷航.卷积神经网络研究综述[J].计算机应用,2016,36(9):2508-2515.Li Yandong,Hao Zongbo,Lei Hang.Review of convolutional neural network[J].Journal of Computer Applications,2016,36(9):2508-2515.(in Chinese)
[10]Li M,Zhang T,Chen Y,et al.Efficient mini-batch training for stochastic optimization[C].Acm Sigkdd International Conference on Knowledge Discovery&Data Mining.ACM,2014:661-670.
[11]Krizhevsky A,Sutskever I,Hinton G E.ImageNet classification with deep convolutional neural networks[C].International Conference on Neural Information Processing Systems.Curran Associates Inc.2012:1097-1105.
[12]The Case Western Reserve University Bearing Data Center Seeded Fault Test Data[EB/OL].[2018-10-10].http:∥csegroups.case.edu/bearingdatacenter/pages/download-data-file.
[13]张伟.基于卷积神经网络的轴承故障诊断算法研究[D].哈尔滨:哈尔滨工业大学,2017.