旋转部件是否发生局部故障，关键是判断其振动信号在空间上是否出现周期性冲击以及周期大小。卷积神经网络(CNN)善于挖掘数据空间上的局部重要的信息特征，具有“端对端”的优势，从而克服了人工提取特征的缺陷;由于振动信号在时间维度上也蕴含着丰富的信息，而长短时记忆网络(LSTM)善于从动态变化的序列数据中学习到时间上的关联性;门控递归单元（GRU）属于LSTM的变种，但相对于LSTM结构更加简洁，参数的数量更少，因此将CNN的空间处理能力和GRU时序处理能力的优势结合，提出一种双通道特征融合CNN-GRU齿轮箱故障诊断方法;即采用并列式结构令CNN与GRU双通道同时提取齿轮箱原始振动信号的故障特征，然后将双通道提取的特征向量合并成一个融合特征向量，输入到SoftMax进行故障分类。该方法可以直接从原始振动信号自适应提取到空间和时序的融合特征，实现了“端对端”的故障诊断。用齿轮实测数据和西储大学轴承数据进行验证，试验结果表明，所提方法识别准确率较高，具有实用性和可行性。
Abstract
Whether a rotating part has local fault or not, the key is to judge if its vibration signal has periodic impact in space and period size. Convolutional neural network (CNN) is good at mining local important information features in data space, and has the advantage of "end-to-end" to overcome the defect of manual feature extraction. As vibration signal also contains rich information in time domain, the long-short term memory (LSTM) network is good at learning temporal relevance from dynamic sequence data. The gate recurrent unit (GRU) is a variant of LSTM, but compared with LSTM, its structure is more concise and the number of parameters is less. Here, combining advantages of CNN’s spatial processing ability and GRU’s temporal processing ability, a dual-channel feature fusion CNN-GRU gearbox fault diagnosis method was proposed. A parallel structure was adopted to make CNN and GRU extract fault features of the original vibration signal of gearbox simultaneously, and then feature vectors extracted from the two channels were combined into a fused feature vector, it was input into the software SoftMax for fault classification. It was shown that the proposed method can adaptively extract fused features of space and time sequencies directly from the original vibration signal to realize the "end-to-end" fault diagnosis. The test results showed that the proposed method has a higher recognition accuracy, and a certain practicability and feasibility.
关键词
齿轮箱 /
卷积神经网络(CNN) /
门控递归单元(GRU) /
故障诊断
{{custom_keyword}} /
Key words
gearbox /
convolutional neural network (CNN) /
gated recurrent unit (GRU) /
fault diagnosis
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］SHAO H D, JIANG H K, LIN Y, et al. A novel method for intelligent fault diagnosis of rolling bearings using ensemble deep auto-encoders［J］. Mechanical Systems and Signal Processing, 2018, 102: 278-297.
［2］HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis［J］. Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences, 1998, 454(1971): 903-995.
［3］WANG W J, MCFADDEN P D. Application of wavelets to gearbox vibration signals for fault detection［J］. Journal of Sound and Vibration, 1996, 192(5): 927-939.
［4］李恒，张氢，秦仙蓉，等. 基于短时傅里叶变换和卷积神经网络的轴承故障诊断方法［J］. 振动与冲击, 2018, 37(19): 124-131.
LI Heng，ZHANG Qin，QIN Xianrong, et al. Fault diagnosis method for rolling bearings based on short-time Fourier transform and convolution neural network［J］. Journal of Vibration and Shock, 2018, 37(19): 124-131.
［5］吴守军，冯辅周，吴春志，等. 基于VMD-DE的坦克行星变速箱故障诊断方法研究［J］. 振动与冲击, 2020, 39(10): 170-179.
WU Shoujun，FENG Fuzhou，WU Chunzhi，et al. Research on fault diagnosis method of tank planetary gearbox based on VMD-DE［J］. Journal of Vibration and Shock, 2020, 39(10): 170-179.
［6］HINTON G E, SALAKHUTDINOV R R. Reducing the dimensionality of data with neural networks［J］. Science, 2006, 313(5786): 504-507.
［7］JANSSENS O, SLAVKOVIKJ V, VERVISCH B, et al. Convolutional neural network based fault detection for rotating machinery［J］. Journal of Sound and Vibration, 2016, 377: 331-345.
［8］朱丹宸，张永祥，潘洋洋，等. 基于多传感器信号和卷积神经网络的滚动轴承故障诊断［J］. 振动与冲击, 2020, 39(4): 172-178.
ZHU Danchen，ZHANG Yongxiang，PAN Yangyang，et al. Fault diagnosis for rolling element bearings based on multisensor signals and CNN［J］. Journal of Vibration and Shock, 2020, 39(4): 172-178.
［9］EREN L. Bearing fault detection by one-dimensional convolutional neural networks［J］. Mathematical Problems in Engineering, 2017, 2017: 1-9.
［10］HOCHREITER S, SCHMIDHUBER J. Long short-term memory［J］. Neural Computation, 1997, 9(8): 1735-1780.
［11］LI S, ZHAO Y, DING M. Mobile robot motor bearing fault detection and classification on discrete wavelet transform and LSTM network［J］. Journal of Mechanics in Medicine and Biology, 2018, 18(8): 1840034.
［12］HINCHI A Z, TKIOUAT M. Rolling element bearing remaining useful life estimation based on a convolutional long-short-term memory network［J］. Procedia Computer Science, 2018, 127: 123-132.
［13］宫文峰，陈辉，张泽辉，等. 基于改进卷积神经网络的滚动轴承智能故障诊断研究［J］. 振动工程学报, 2020, 33(2): 400-413.
GONG Wenfeng，CHEN Hui，ZHANG Zehui，et al. Intelligent fault diagnosis for rolling bearing based on improved convolutional neural network［J］. Journal of Vibration Engineering, 2020, 33(2): 400-413.
［14］姜洪权，贺帅，高建民，等. 一种改进卷积神经网络模型的焊缝缺陷识别方法［J］. 机械工程学报, 2020, 56(8): 235-242.
JIANG Hongquan，HE Shuai，GAO Jianmin，et al. An improved convolutional neural network for weld defect recognition［J］. Journal of Mechanical Engineering, 2020, 56(8): 235-242.
［15］JOZEFOWICZ R, ZAREMBA W, SUTSKEVER I. An empirical exploration of recurrent network architectures［J］. International Conference on Machine Learning, 2015,37: 2342-2350.
［16］HE J, YANG S X, PAPATHEOU E, et al. Investigation of a multi-sensor data fusion technique for the fault diagnosis of gearboxes［J］. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 2019, 233(13): 4764-4775.
［17］何俊. 齿轮箱振动特性分析与智能故障诊断方法研究［D］. 杭州:浙江大学, 2018.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(2429 KB)
