为精细化表征风机基础螺栓松动状态特征，实现对风机基础螺栓松动的智能检测，提出多尺度一维卷积神经网络的风机基础螺栓松动智能检测方法。首先，以风机运行时振动时域信号作为多尺度一维卷积神经网络的输入，摆脱对信号处理和专业知识的依赖，并最大程度保留原始信号特征；然后，通过交替的多尺度卷积层和池化层对时域信号特征进行学习，其中多尺度卷积层设置不同尺度的卷积核进行卷积运算，避免单一尺度卷积核对不同精细度特征的忽略，增强网络对特征的表达能力，实现对时域信号特征精细化分布式表征；最后，在特征输出层后添加Softmax多分类器，利用反向传播（Backpropagation,BP）逐层微调结构参数建立特征空间到松动状态空间的映射，输出风机基础螺栓松动检测结果。所提方法将松动特征自动学习与松动识别融为一体，实现了风机基础螺栓松动智能检测。通过在稳定转速和变转速下对风机基础螺栓松动检测实验，证明了所提方法的可行性和有效性。
关键词：风机基础螺栓；松动状态；智能检测；多尺度一维卷积神经网络；精细化表征
Abstract
To refine the looseness state featuresof bolts offan foundation and realizelooseness intelligent detection of bolts offan foundation, a multi-scale one-dimensional convolution neural networklooseness intelligent detection methodforbolts of fan foundation is proposed.Firstly, the vibration time-domain signal is used as the input of multi-scale one-dimensional convolution neural network, which can get rid of the dependence on signal processing and professional knowledge, and retain the original signal features to the greatest extent;Then, the time-domain signal features are learned by alternate multi-scale convolution layer and pooling layer,the multi-scale convolution layer sets convolution kernels of different scales for convolution operation to avoid single-scale convolution kernels ignoring different fineness featuresandincrease the ability to represent features,realizing refined distributed representation of time-domain signal features;Finally,softmax multiple classifiers are added to the end of the feature output layer, the mapping from feature space to loosenessstate space is established by using back propagation to fine tune the structure parameters layer by layer, and looseness detection results for bolts of fan foundationare output.The proposed method integrates the automatic learning of loosenessfeatures with the recognition of looseness, and realize looseness intelligent detection for bolts of fan foundation.The feasibility and effectiveness of the proposed method are proved by looseness detection experiment for bolts offan foundation under the condition of stable speed and variable speed.
Key words:bolts offan foundation;looseness state;intelligent detection; multi-scale one-dimensional convolutional neural network; refined representation
关键词
风机基础螺栓 /
松动状态 /
智能检测 /
多尺度一维卷积神经网络 /
精细化表征
{{custom_keyword}} /
Key words
bolts offan foundation;looseness state;intelligent detection /
multi-scale one-dimensional convolutional neural network /
refined representation
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 《中国公路学报》编辑部. 中国隧道工程学术研究综述：2015[J].中国公路学报,2015,28(5):1-65.
Editorial Department of China Journal of Highway and Transport.Review on China’s tunnel engineering academic research:2015[J]. China Journal of Highway and Transport, 2015,28(5):1-65.
[2] 韩坤林.基于振动分析的公路隧道悬挂风机基础健康性监测方法[J].公路隧道,2016(1):45-48.
HAN Kunlin. Health monitoring method for suspension fan foundation of highway tunnel based on vibration analysis[J]. Highway Tunnel, 2016 (1): 45-48.
[3] 陈仁祥,周君,杨黎霞,等.振动敏感特征与流形学习的风机基座螺栓松动程度诊断[J].计算机集成制造系统,2017,23(12):2613-2621.
CHEN Renxiang, ZHOU Jun, YANG Lixia,et al. Diagnosis of vibration sensitive characteristics and looseness degree of fan base bolts by manifold learning[J]. Computer Integrated Manufacturing System, 2017,23 (12): 2613-2621.
[4] 周飞燕,金林鹏,董军.卷积神经网络研究综述[J].计算机学报,2017,40(6):1229-1251.
ZHOU Feiyan, JIN Linpeng, DONG Jun. Review of convolution neural network research[J]. Journal of Computer Science, 2017,40 (6): 1229-1251.
[5] ZHANG W, PENG G L, LI C H.Bearings fault diagnosis based on convolution neural network with 2D representation of vibration signals as input[J]. MATEC Web of Conferences,2017,95(8):13001.
[6] WEN L, LI X, GAO L,et al. A new convolutional neural network based data-driven fault diagnosis method[J]. IEEE Transactions on Industrial Electronics, 2018,65(7):5990-5998.
[7] YANG Z B, JIA M P. Ga-1dlcnn method and its application in bearing fault diagnosis[J].Journal of Southeast University,2019,35(1):36-42.
[8] 安晶,艾萍,徐森,等.一种基于一维卷积神经网络的旋转机械智能故障诊断方法[J].南京大学学报(自然科学),2019,55(1):133-142.
AN Jing, AI Ping, XU Sen, et al.An intelligent fault diagnosis method for rotating machinery based on one-dimensional convolutional neural network [J]. Journal of Nanjing University (Natural Science), 2019,55 (1): 133-142.
[9] 周奇才,刘星辰,赵炯,等.旋转机械一维深度卷积神经网络故障诊断研究[J].振动与冲击,2018,37(23):31-37.
ZHOU Qicai, LIU Xingchen, ZHAO Jiong, et al. Research on fault diagnosis of one dimensional deep convolution neural network of rotating machinery[J]. Journal of Vibration and Shock, 2018,37 (23): 31-37.
[10] SZEGEDY C, LIU W,JIA Y, et al. Going deeper with convolutions[C]//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition. Boston:IEEE, 2015.
[11] 郭晨,简涛,徐从安, 等. 基于深度多尺度一维卷积神经网络的雷达舰船目标识别[J]. 电子与信息学报, 2019, 41(6): 1302-1309.
GUO Chen, JIAN Tao, XU Cong＇an, et al. Radar HRRP target recognition based on deep multi-scale 1d convolutional neural network[J]. Journal of Electronics and Information Technology, 2019, 41(6): 1302-1309.
[12] 吴俊,管鲁阳,鲍明,等.基于多尺度一维卷积神经网络的光纤振动事件识别[J].光电工程,2019,46(5):79-86.
WU Jun, GUAN Luyang, BAO Ming, et al.Optical fiber vibration event recognition based on multi-scale one-dimensional convolution neural network[J]. Optoelectronic Engineering, 2019,46 (5): 79-86.
[13] JING L,ZHAO M,LI P, et al. A convolutional neural network based feature learning and fault diagnosis method for the condition monitoring of gearbox[J].Measurement,2017,111:1-10．
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(1743 KB)
