Application of WPD and SVM-PSO in Online Monitoring of Micro Milling Tool Wear
-
摘要: 为提高微铣刀磨损状态的预测精度和计算效率,本文提出了一种基于小波包分解(Wavelet packet decomposition, WPD)和支持向量机-粒子群优化(Support vector machine-particle swarm optimization, SVM-PSO)的微铣刀磨损在线监测方法。首先根据刀具使用时长和磨损程度将微铣刀磨损分为初始磨损、轻度磨损、中度磨损、重度磨损和刀具失效5种状态;接着对采集到的振动信号进行WPD变换,提取小波包关键节点的能量比和小波包系数峭度作为磨损特征,并分析了不同切削参数对这2个特征的影响;最后利用SVM-PSO模型进行微铣刀磨损状态分类与预测。研究结果表明,和网格搜索法相比,本文提出的微铣刀磨损在线监测方法在计算精度和效率方面具有综合优势,可以为其它刀具磨损监测提供必要的理论基础和实践指导。Abstract: In order to improve the prediction accuracy and calculation efficiency of tool wear state in the micro milling, an online monitoring method of tool wear in the micro milling based on the wavelet packet decomposition (WPD) and support vector machine-particle swarm optimization (SVM-PSO) is put forward. Firstly, the wear of tool in the micro milling can be divided into the five states: initial wear, light wear, medium wear, heavy wear and tool failure. Secondly, the collected vibration signals are transformed by using WPD, and the energy ratio and kurtosis of key nodes of wavelet packet are extracted as the wear features, and the influence of the different cutting parameters on the two features is analyzed. Finally, SVM-PSO model is used to classify and predict the wear state of tool in the micro milling. The results show that, comparing with the grid search method, the online wear monitoring method of tool in the micro milling proposed in this paper has comprehensive advantages in the calculation accuracy and efficiency, and can provide the necessary basis and guidance for monitoring the other tool wear.
-
表 1 4种核函数及其数学表达式
核函数类型 数学表达式 线性 径向基函数 多项式 Sigmoid 表 2 模具钢NAK80的微铣削切削参数
序号 主轴转速/ (r·min-1) 切削深度/mm 进给速度/ (mm·r-1) 采样频率/Hz 1 16 000 0.010 0.003 5 000 2 16 000 0.015 0.004 5 000 3 16 000 0.020 0.005 5 000 4 16 000 0.025 0.006 5 000 5 16 000 0.030 0.007 5 000 6 16 000 0.010 0.005 5 000 7 16 000 0.015 0.006 5 000 8 18 000 0.010 0.003 5 000 9 18 000 0.015 0.004 5 000 10 18 000 0.020 0.005 5 000 11 18 000 0.025 0.006 5 000 12 18 000 0.030 0.007 5 000 13 18 000 0.010 0.006 5 000 14 18 000 0.015 0.007 5 000 15 20 000 0.010 0.003 5 000 16 20 000 0.015 0.004 5 000 17 20 000 0.020 0.005 5 000 18 20 000 0.025 0.006 5 000 19 20 000 0.030 0.007 5 000 20 20 000 0.010 0.005 5 000 21 20 000 0.015 0.006 5 000 22 20 000 0.020 0.004 5 000 23 21 000 0.010 0.003 5 000 24 21 000 0.015 0.004 5 000 25 21 000 0.020 0.005 5 000 26 21 000 0.025 0.006 5 000 27 21 000 0.030 0.007 5 000 28 21 000 0.010 0.005 5 000 29 21 000 0.015 0.006 5 000 30 22 000 0.010 0.003 5 000 31 22 000 0.015 0.004 5 000 32 22 000 0.020 0.005 5 000 33 22 000 0.025 0.006 5 000 34 22 000 0.030 0.007 5 000 35 22 000 0.010 0.005 5 000 36 22 000 0.015 0.006 5 000 表 3 PSO的参数设置
r1 r2 c1 c2 w 0.5 0.5 2 2 1.2 -
[1] ZHU K P, YU X L. The monitoring of micro milling tool wear conditions by wear area estimation[J]. Mechanical Systems and Signal Processing, 2017, 93: 80-91 doi: 10.1016/j.ymssp.2017.02.004 [2] KURADA S, BRADLEY C. A machine vision system for tool wear assessment[J]. Tribology International, 1997, 30(4): 295-304 doi: 10.1016/S0301-679X(96)00058-8 [3] GARCÍA-ORDÁS M T, ALEGRE E, GONZÂLEZ-CASTRO V, et al. A computer vision approach to analyze and classify tool wear level in milling processes using shape descriptors and machine learning techniques[J]. The International Journal of Advanced Manufacturing Technology, 2017, 90(5-8): 1947-1961 doi: 10.1007/s00170-016-9541-0 [4] MOLDOVAN O G, DZITAC S, MOGA I, et al. Tool-wear analysis using image processing of the tool flank[J]. Symmetry, 2017, 9(12): 296 doi: 10.3390/sym9120296 [5] D'ADDONA D M, ULLAH A M M S, MATARAZZO D. Tool-wear prediction and pattern-recognition using artificial neural network and DNA-based computing[J]. Journal of Intelligent Manufacturing, 2017, 28(6): 1285-1301 doi: 10.1007/s10845-015-1155-0 [6] ONG P, LEE W K, LAU R J H. Tool condition monitoring in CNC end milling using wavelet neural network based on machine vision[J]. The International Journal of Advanced Manufacturing Technology, 2019, 104(1): 1369-1379 [7] LEE W K, RATNAM M M, AHMAD Z A. Detection of chipping in ceramic cutting inserts from workpiece profile during turning using fast Fourier transform (FFT) and continuous wavelet transform (CWT)[J]. Precision Engineering, 2017, 47: 406-423 doi: 10.1016/j.precisioneng.2016.09.014 [8] LI X L. Development of current sensor for cutting force measurement in turning[J]. IEEE Transactions on Instrumentation and Measurement, 2005, 54(1): 289-296 doi: 10.1109/TIM.2004.840225 [9] DIMLA SR D E, LISTER P M. On-line metal cutting tool condition monitoring. : I: force and vibration analyses[J]. International Journal of Machine Tools and Manufacture, 2000, 40(5): 739-768 doi: 10.1016/S0890-6955(99)00084-X [10] ZHU K P, VOGEL-HEUSER B. Sparse representation and its applications in micro-milling condition monitoring: noise separation and tool condition monitoring[J]. The International Journal of Advanced Manufacturing Technology, 2014, 70(1-4): 185-199 doi: 10.1007/s00170-013-5258-5 [11] PRAKASH, KANTHABABU M. In-process tool condition monitoring using acoustic emission sensor in microendmilling[J]. Machining Science and Technology, 2013, 17(2): 209-227 doi: 10.1080/10910344.2013.780541 [12] DURO J A, PADGET J A, BOWEN C R, et al. Multi-sensor data fusion framework for CNC machining monitoring[J]. Mechanical Systems and Signal Processing, 2016, 66-67: 505-520 doi: 10.1016/j.ymssp.2015.04.019 [13] MALEKIAN M, PARK S S, JUN M B G. Tool wear monitoring of micro-milling operations[J]. Journal of Materials Processing Technology, 2009, 209(10): 4903-4914 doi: 10.1016/j.jmatprotec.2009.01.013 [14] WANG L, LIANG M. Chatter detection based on probability distribution of wavelet modulus Maxima[J]. Robotics and Computer-Integrated Manufacturing, 2009, 25(6): 989-998 doi: 10.1016/j.rcim.2009.04.011 [15] ZHOU Y Q, CHANG S H, WU S G, et al. FFT-ApEn analysis for the vibration signal of a rotating motor[J]. International Journal of Acoustics and Vibration, 2018, 23(2): 203-207 [16] DZIEDZIECH K, JABLONSKI A, DWORAKOWSKI Z. A novel method for speed recovery from vibration signal under highly non-stationary conditions[J]. Measurement, 2018, 128: 13-22 doi: 10.1016/j.measurement.2018.06.024 [17] ZHEN G, LI X B, YU T Y, et al. Vibration signal EMD filter detection method for blast furnace opening machine[M]//JIA Y M, DU J P, ZHANG W C. Proceedings of 2018 Chinese Intelligent Systems Conference. Singapore: Springer, 2019: 173-180 [18] YAO Z H, MEI D Q, CHEN Z C. On-line chatter detection and identification based on wavelet and support vector machine[J]. Journal of Materials Processing Technology, 2010, 210(5): 713-719 doi: 10.1016/j.jmatprotec.2009.11.007 [19] LIU J, HU Y M, WU B, et al. A hybrid health condition monitoring method in milling operations[J]. The International Journal of Advanced Manufacturing Technology, 2017, 92(5): 2069-2080 [20] 郭庆丰, 王成栋, 刘佩森. 时域指标和峭度分析法在滚动轴承故障诊断中的应用[J]. 机械传动, 2016, 40(11): 172-175 https://www.cnki.com.cn/Article/CJFDTOTAL-JXCD201611049.htmGUO Q F, WANG C D, LIU P S. Application of time domain index and kurtosis analysis method in the fault diagnosis of rolling bearing[J]. Journal of Mechanical Transmission, 2016, 40(11): 172-175 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JXCD201611049.htm