深度学习与多信号融合在铣刀磨损状态识别中的研究

穆殿方; 刘献礼; 岳彩旭; Steven Y. LIANG; 陈志涛; 李恒帅; 徐梦迪

doi:10.13433/j.cnki.1003-8728.20200209

深度学习与多信号融合在铣刀磨损状态识别中的研究

doi: 10.13433/j.cnki.1003-8728.20200209

1.
哈尔滨理工大学机械动力工程学院, 哈尔滨 150080
2.
佐治亚理工学院乔治·W·伍德拉夫机械工程学院, 美国亚特兰大 30332

基金项目:

国家自然科学基金国际(地区)合作与交流重点项目 51720105009

黑龙江省自然科学基金优秀青年项目 YQ2019E029

详细信息

作者简介:
穆殿方(1989-), 硕士研究生, 研究方向为信号处理、数据融合, 1297054155@qq.com

通讯作者:
刘献礼, 教授, 博士生导师, xlliu@hrbust.edu

中图分类号: TG502; TP306⁺.2
计量
- 文章访问数: 216
- HTML全文浏览量: 63
- PDF下载量: 32
- 被引次数: 0
出版历程
- 收稿日期: 2020-04-13
- 刊出日期: 2021-10-01

Study on Wear State Recognition of Milling Cutter via Deep Learning and Multi-signal Fusion

1.
School of Mechanical and Power Engineering, Harbin University of Science and Technology, Harbin 150080, China
2.
The George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta 30332, USA

摘要

摘要: 为精确地识别刀具磨损状态，提出了一种深度学习与多信号融合相结合的识别方法。以自编码网络为基础，构建了堆叠稀疏自编码网络。采集铣刀不同磨损状态下的力信号、振动信号及声发射信号，并对上述信号进行小波包分解以便获取能够表征铣刀磨损的时频域特征。利用无监督学习和有监督学习对堆叠稀疏自编码网络进行训练，建立了深度学习的铣刀磨损状态识别模型。研究结果表明，多信号融合的深度学习模型对铣刀磨损状态识别准确率达到94.44%。
- 刀具磨损 /
- 状态识别 /
- 深度学习 /
- 多信号融合 /
- 堆叠稀疏自编码网络
Abstract: The cutting tool is one of the most active factors in the machining, its status directly affects the surface quality of the workpiece. To accurately identify tool wear, a recognition method combining deep learning with multi-signal fusion was proposed. Based on the autoencoder network, the stacked sparse autoencoder network was constructed. The force signal, vibration signal and acoustic emission signal were collected under different wear condition of milling cutter, and the wavelet packet decomposition was carried out to obtain the time-frequency characteristics of milling cutter wear. The stacked sparse autoencoder network was trained by using the unsupervised learning and supervised learning, a recognition model for wear state of milling tool based on the deep learning was established. The results show that the deep learning model combing with the multi-signal fusion has an accuracy rate of 94.44% for identifying the wear state of milling cutter, the results lay a foundation for controlling the optimization of milling process.
- tool wear /
- state recognition /
- deep learning /
- multi-signal fusion /
- stacked sparse autoencoder network

HTML全文

图 1 自编码器网络结构图

下载: 全尺寸图片幻灯片

图 2 堆叠稀疏自编码网络

下载: 全尺寸图片幻灯片

图 3 有监督的网络结构

下载: 全尺寸图片幻灯片

图 4 深度学习的铣刀磨损状态识别流程

下载: 全尺寸图片幻灯片

图 5 信号采集现场图

下载: 全尺寸图片幻灯片

图 6 不同切削参数的铣刀磨损曲线

下载: 全尺寸图片幻灯片

图 7 铣刀不同磨损状态下的测量图

下载: 全尺寸图片幻灯片

图 8 信号截断处理

下载: 全尺寸图片幻灯片

图 9 力信号不同频段的能量值

下载: 全尺寸图片幻灯片

图 10 振动信号不同频段的能量值

下载: 全尺寸图片幻灯片

图 11 声发射信号不同频段的能量值

下载: 全尺寸图片幻灯片

图 12 训练误差曲线

下载: 全尺寸图片幻灯片

图 13 不同隐藏层数下SSAE的测试准确度

下载: 全尺寸图片幻灯片

图 14 刀具类型分类模型的混淆矩阵实例图

下载: 全尺寸图片幻灯片

图 15 单信号特征对铣刀磨损状态识别结果

下载: 全尺寸图片幻灯片

图 16 不同的算法模型对铣刀磨损状态识别精度

下载: 全尺寸图片幻灯片

图 17 深度学习识别结果

下载: 全尺寸图片幻灯片

表 1 铣刀信息

刀具材质	涂层材料	直径直径/mm	刀具前角/(°)	刀具后角/(°)
硬质合金	无涂层	10	8	9

下载: 导出CSV

表 2 切削参数

实验序号	主轴转速/(rad·min^-1)	切削深度/mm	切削宽度/mm	进给速度/(mm·min^-1)
1	2 000	6	0.3	400
2	2 000	3	0.3	400

下载: 导出CSV

表 3 有监督学习样本标签

磨损状态分类	磨损值的范围/mm	标签
初期磨损	0~0.1	[1 0 0]
正常磨损	0.1~0.201	[0 1 0]
严重磨损	＞0.201	[0 0 1]

下载: 导出CSV

参考文献(20)

[1]	DU J B, YUE C X, LIU X L, et al. Transient temperature field model of wear land on the flank of end mills: a focus on time-varying heat intensity and time-varying heat distribution ratio[J]. Applied Sciences, 2019, 9(8): 1698 doi: 10.3390/app9081698
[2]	CAGGIANO A. Tool wear prediction in Ti-6Al-4V machining through multiple sensor monitoring and PCA features pattern recognition[J]. Sensors, 2018, 18(3): 823 doi: 10.3390/s18030823
[3]	刘献礼, 刘强, 岳彩旭, 等. 切削过程中的智能技术[J]. 机械工程学报, 2018, 54(16): 45-61 https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201816005.htm LIU X L, LIU Q, YUE C X, et al. Intelligent machining technology in cutting process[J]. Journal of Mechanical Engineering, 2018, 54(16): 45-61 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201816005.htm
[4]	WANG C D, BAO Z L, ZHANG P P, et al. Tool wear evaluation under minimum quantity lubrication by clustering energy of acoustic emission burst signals[J]. Measurement, 2019, 138: 256-265 doi: 10.1016/j.measurement.2019.02.004
[5]	SCHWENZER M, MIURA K, BERGS T. Machine learning for tool wear classification in milling based on force and current sensors[C]//2019 4th International Conference on Mechanical, Manufacturing, Modeling and Mechatronics (IC4M 2019) & 2019 4th International Conference on Design Engineering and Science (ICDES 2019). IOP Publishing: London, 2019: 69-77
[6]	COPPEL R, ABELLAN-NEBOT J V, SILLER H R, et al. Adaptive control optimization in micro-milling of hardened steels-evaluation of optimization approaches[J]. The International Journal of Advanced Manufacturing Technology, 2016, 84(9-12): 2219-2238 doi: 10.1007/s00170-015-7807-6
[7]	WU J, SU Y, CHENG Y, et al. Multi-sensor information fusion for remaining useful life prediction of machining tools by adaptive network based fuzzy inference system[J]. Applied Soft Computing, 2018, 68: 13-23 doi: 10.1016/j.asoc.2018.03.043
[8]	ZHANG K F, YUAN H Q, NIE P. A method for tool condition monitoring based on sensor fusion[J]. Journal of Intelligent Manufacturing, 2015, 26(5): 1011-1026 doi: 10.1007/s10845-015-1112-y
[9]	LIU C Q, LI Y G, ZHOU G Y, et al. A sensor fusion and support vector machine based approach for recognition of complex machining conditions[J]. Journal of Intelligent Manufacturing, 2018, 29(8): 1739-1752 doi: 10.1007/s10845-016-1209-y
[10]	陈刚, 焦黎, 颜培, 等. 基于多传感器数据融合的刀具磨损状态监测研究[J]. 新技术新工艺, 2017(11): 23-28 https://www.cnki.com.cn/Article/CJFDTOTAL-XJXG201711007.htm CHEN G, JIAO L, YAN P, et al. Research on tool wear condition monitoring based on multi-sensor data fusion[J]. New Technology & New Process, 2017(11): 23-28 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XJXG201711007.htm
[11]	HINTON G E, OSINDERO S, TEH Y W. A fast learning algorithm for deep belief nets[J]. Neural Computation, 2006, 18(7): 1527-1554 doi: 10.1162/neco.2006.18.7.1527
[12]	TERRAZAS G, MARTÍNEZ-ARELLANO G, BENARDOS P, et al. Online tool wear classification during dry machining using real time cutting force measurements and a CNN approach[J]. Journal of Manufacturing and Materials Processing, 2018, 2(4): 72 doi: 10.3390/jmmp2040072
[13]	林杨, 高思煜, 刘同舜, 等. 基于深度学习的高速铣削刀具磨损状态预测方法[J]. 机械与电子, 2017, 35(7): 12-17 doi: 10.3969/j.issn.1001-2257.2017.07.003 LIN Y, GAO S Y, LIU T S, et al. A deep learning-based method for tool wear state prediction in high speed milling[J]. Machinery & Electronics, 2017, 35(7): 12-17 (in Chinese) doi: 10.3969/j.issn.1001-2257.2017.07.003
[14]	MARTÍNEZ-ARELLANO G, TERRAZAS G, RATCHEV S. Tool wear classification using time series imaging and deep learning[J]. International Journal of Advanced Manufacturing Technology, 2019, 104(9): 3647-3662
[15]	李晴晴, 侯瑞春, 丁香乾. 基于改进堆叠自编码器的滚动轴承故障诊断[J]. 计算机工程与设计, 2019, 40(7): 2064-2070 https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201907043.htm LI Q Q, HOU R C, DING X Q. Roller bearing fault diagnosis based on improved stacked auto-encoder[J]. Computer Engineering and Design, 2019, 40(7): 2064-2070 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201907043.htm
[16]	戴稳, 张超勇, 孟磊磊, 等. 基于深度学习与特征后处理的支持向量机铣刀磨损预测模型[J/OL]. 计算机集成制造系统, (2019-03-04)[2020-02-22]. http://kns.cnki.net/kcms/detail/11.5946.TP.20190301.1140.012.html DAI W, ZHANG C Y, MENG L L, et al. Wear prediction model of SVM milling cutter based on deep learning and feature post-processing[J/OL]. Computer Integrated Manufacturing System, (2019-03-04)[2020-02-22]. http://kns.cnki.net/kcms/detail/11.5946.TP.20190301.1140.012.html (in Chinese)
[17]	FU Y, ZHANG Y, QIAO H Y, et al. Analysis of feature extracting ability for cutting state monitoring using deep belief networks[J]. Procedia CIRP, 2015, 31: 29-34 doi: 10.1016/j.procir.2015.03.016
[18]	王国锋, 董毅, 杨凯, 等. 基于深度学习与粒子滤波的刀具寿命预测[J]. 天津大学学报, 2019, 52(11): 1109-1116 https://www.cnki.com.cn/Article/CJFDTOTAL-TJDX201911001.htm WANG G F, DONG Y, YANG K, et al. Tool life prediction based on deep learning and particle filtering[J]. Journal of Tianjin University, 2019, 52(11): 1109-1116 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-TJDX201911001.htm
[19]	王丽华, 杨家巍, 张永宏, 等. 基于堆叠降噪自编码的刀具磨损状态识别[J]. 中国机械工程, 2018, 29(17): 2038-2045 doi: 10.3969/j.issn.1004-132X.2018.17.004 WANG L H, YANG J W, ZHANG Y H, et al. Tool wear condition recognition based on SDAE[J]. China Mechanical Engineering, 2018, 29(17): 2038-2045 (in Chinese) doi: 10.3969/j.issn.1004-132X.2018.17.004
[20]	李宏坤, 郝佰田, 代月帮, 等. 基于压缩感知和加噪堆栈稀疏自编码器的铣刀磨损程度识别方法研究[J]. 机械工程学报, 2019, 55(14): 1-10 https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201914001.htm LI H K, HAO B T, DAI Y B, et al. Wear status recognition for milling cutter based on compressed sensing and noise stacking sparse auto-encoder[J]. Journal of Mechanical Engineering, 2019, 55(14): 1-10(in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201914001.htm