多传感器信息深度融合的谐波减速器健康状态评估

PDF(1688 KB)

振动与冲击 ›› 2022, Vol. 41 ›› Issue (7) : 139-144.

论文

陈仁祥1,张勇1，胡小林2，杨黎霞1，陈才3，谢文举1

作者信息 +

Health state evaluation of harmonic reducer based on multi-sensor information deep fusion

CHEN Renxiang1, ZHANG Yong1, HU Xiaolin2, YANG Lixia1, CHEN Cai3, XIE Wenju1

Author information +

文章历史 +

摘要

工业机器人谐波减速器工况循环往复，仅依靠单一传感器难以刻画其运行状态全貌且会导致健康状态评估结果不确定性高。为此，提出了多传感器信息深度融合的谐波减速器健康状态评估方法。首先对谐波减速器振动信号进行连续小波变换，构造出时频图以描述其运行状态特征；再运用基于小波变换的图像融合方法将多个传感器的时频信息进行融合以全面刻画谐波减速器运行状态。最后利用卷积神经网络对融合后的时频图像进行自动学习获得能准确表征谐波减速器健康状态的深度特征，并通过在卷积神经网络最后添加全连接层实现健康状态评估。通过对不同健康状态以及不同工作节拍下谐波减速器进行健康状态评估实验，证明了所提方法的可行性与有效性，并具有较好的泛化能力和稳健性。

Abstract

The operating condition of industrial robot harmonic reducer is repeated circularly, so it is difficult to describe its operating state only by a single sensor, and it will lead to high uncertainty of health state assessment results. To this end, a method of evaluating the health state of harmonic reducer with multi-sensor information fusion deeply is proposed. Firstly, continuous wavelet transform is applied to the vibration signal of harmonic reducer, and the time-frequency diagram is constructed to describe its operating characteristics. Then the image fusion based on wavelet transform is used to fuse the time-frequency information of multiple sensors to describe the running state of harmonic reducer comprehensively. Finally, the fused time-frequency diagram is automatically learned by using convolutional neural network to obtain the depth characteristics that can accurately represent the health state of harmonic reducer, and the health state assessment is realized by adding full connection layer at the end of the convolutional neural network. The feasibility and effectiveness of the proposed method are proved by the experiment of evaluating the health state of the harmonic reducer under different health state and different working time, and it has good generalization ability and robustness.

导出引用

陈仁祥1,张勇1，胡小林2，杨黎霞1，陈才3，谢文举1. 多传感器信息深度融合的谐波减速器健康状态评估[J]. 振动与冲击, 2022, 41(7): 139-144

CHEN Renxiang1, ZHANG Yong1, HU Xiaolin2, YANG Lixia1, CHEN Cai3, XIE Wenju1. Health state evaluation of harmonic reducer based on multi-sensor information deep fusion[J]. Journal of Vibration and Shock, 2022, 41(7): 139-144

参考文献

[1] 王田苗, 陶永. 我国工业机器人技术现状与产业化发展战略[J]. 机械工程学报, 2014, 50(9):1-13.
WANG Tian-miao, TAO Yong. Current situation and industrialization development strategy of industrial robotics in China[J]. Journal of Mechanical Engineering, 2014, 50(9):1-13.
[2] 陈仁祥, 张勇, 杨黎霞, 等. 基于整周期数据和卷积神经网络的谐波减速器健康状态评估[J]. 仪器仪表学报, 2020, 41(02):245-252.
CHEN Ren-xiang, ZHANG Yong, YANG Li-xia, et al. Health status assessment of harmonic speed reducer based on integer-cycle data and convolutional neural network[J]. Chinese Journal of Scientific Instrument, 2020, 41(02):245-252.
[3] 张明, 江志农. 基于多源信息融合的往复式压缩机故障诊断方法[J]. 机械工程学报, 2017, 53(23):46-52.
ZHANG Ming, JIANG Zhi-nong. Fault diagnosis method of reciprocating compressor based on multi-source information fusion[J]. Journal of Mechanical Engineering, 2017, 53 (23): 46-52.
[4] 王秀青, 侯增广, 曾慧, 等. 基于多传感器信息融合的机器人故障诊断[J]. 上海交通大学学报, 2015, 49(06):793-798.
WANG Xiu-qing, HOU Zeng-guang, ZENG Hui, et al. Robot fault diagnosis based on multi-sensor information fusion[J]. Journal of Shanghai Jiao Tong University, 2015, 49(06):793-798.
[5] CAI Bao-ping, LIU Yong-hong, FAN Qian, et al. Multi-source information fusion based fault diagnosis of ground-source heat pump using Bayesian network[J]. Applied Energy, 2014, 114(2):1-9.
[6] 李松柏, 康子剑, 陶洁. 基于信息融合及堆栈降噪自编码的齿轮故障诊断[J]. 振动与冲击, 2019, 38(05):216-221.
LI Song-bai, KANG Zi-jian, TAO Jie. Gear fault diagnosis based on information fusion and stack noise reduction self-coding[J]. Journal of Vibration and Shock, 2019, 38(05):216-221.
[7] 向丹, 岑健. 基于EMD熵特征融合的滚动轴承故障诊断方法[J]. 航空动力学报, 2015, 30(05):1149-1155.
XIANG Dan, CEN Jian.. Rolling bearing fault diagnosis method based on EMD entropy feature fusion[J]. Journal of Aerospace Power, 2015, 30(05):1149-1155.
[8] Safizadeh M S, Latifi S K. Using multi-sensor data fusion for vibration fault diagnosis of rolling element bearings by accelerometer and load cell[J]. Information Fusion, 2014, 18(4):1-8.
[9] 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(06):1229-1251.
ZHOU Fei-yan, JIN Lin-peng, DONG Jun. Overview of convolutional neural networks[J]. Chinese Journal of Computers, 2017, 40(06):1229-1251.
[10] 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(02):400-413.
GONG Wen-feng, CHEN Hui, ZHANG Ze-hui, et al. Research on intelligent fault diagnosis of rolling bearings based on improved convolutional neural network [J]. Journal of Vibration Engineering, 2020, 33(02):400-413.
[11] 陈仁祥, 黄鑫, 杨黎霞, 等. 基于卷积神经网络和离散小波变换的滚动轴承故障诊断[J]. 振动工程学报, 2018, 31(5):883-891.
CHEN Ren-xiang, HUANG Xin, YANG Li-xia, et al. Fault diagnosis of rolling bearing based on convolutional neural network and discrete wavelet transform[J]. Journal of Vibration Engineering, 2008, 31(5):883-891.
[12] CHEN Zhu-yun, Mauricio Alexandtr, LI Wei-hua, et al. A deep learning method for bearing fault diagnosis based on Cyclic Spectral Coherence and Convolutional Neural Networks [J]. Mechanical Systems and Signal Processing, 2020, 140.
[13] 李宏坤, 刘洪轶, 徐福健, 等. 连续小波最优重构尺度确定方法与故障早期识别[J]. 机械工程学报, 2014, 50(17):69-76.
LI Hong-kun, LIU Hong-yi, XU Fu-jian, et al. Determination method of optimal reconstruction scale of continuous wavelet and early fault identification [J]. Journal of Mechanical Engineering, 2014, 50(17):69-76.
[14] 武治国, 王延杰, 李桂菊. 应用小波变换的自适应脉冲耦合神经网络在图像融合中的应用[J]. 光学精密工程, 2010, 18(003):708-715.
WU Zhi-guo, WANG Yan-jie, LI Gui-ju. Application of Wavelet Transform adaptive Pulse-Coupled Neural Network in Image Fusion [J]. Optical Precision Engineering, 2010, 18(003):708-715.
[15] 晁锐, 张科, 李言俊. 一种基于小波变换的图像融合算法[J]. 电子学报, 2004, 32(5):750-753.
CHAO Rui, ZHANG Ke, LI Yan-jun. An image fusion algorithm based on wavelet transform [J]. Acta Electronica Sinica, 2004, 32(5):750-753.
[16] 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):132- 139.
LI Heng, ZHANG Qing, QIN Xian-rong, et al. Fault diagnosis method of bearing based on short-time Fourier transform and convolutional neural network [J]. Journal of Vibration and Shock, 2018, 37(19):132-139.