-
摘要: 基于稀疏表示理论,提出了一种采用可调品质因子小波变换(TQWT)的滚动轴承故障诊断新方法,分析了包含早期故障成分的原始采集振动信号的特点和早期故障信号的特性,研究了稀疏表示模型在解决故障特征提取问题和故障类型识别问题的应用;运用TQWT将原始信号转换为一组子带小波系数集,研究了利用迭代收缩阈值算法提取出稀疏小波系数的有效性和谱峭度对故障冲击信号敏感的特性,通过计算各子带信号分量的谱峭度,选取包含故障信息明显的子带小波系数,建立了包含稀疏故障信号分量的故障特征提取方法;利用提取出的故障信号稀疏表示分类模型,实现了基于稀疏表示的滚动轴承故障诊断方法。试验结果表明:在凯斯西储数据集上,提出的故障特征提取方法在剔除干扰成分方面有显著效果,提出方法对于4种类型数据的平均诊断准确率为99.83%,对于10种类型数据的平均诊断准确率为97.73%;与只运用TQWT和迭代收缩阈值算法进行故障特征提取的方法相比,故障诊断精度提高了11.60%,算法运行时间减小8%;在QPZZ-Ⅱ旋转机械平台采集到的振动数据集上,提出的方法对于4种类型数据的平均诊断准确率为100%;与传统小波去噪方法相比,准确率提高了35.67%,算法运行时间减小了7.25%。可见,本文提出的方法可以有效解决滚动轴承故障诊断问题。Abstract: Based on the sparse representation theory, a new method of rolling bearing fault diagnosis was proposed using the tunable-Q wavelet transform (TQWT). The characteristics of the original vibration signals and early fault signals containing early fault components were analyzed, and the applications of the sparse representation model to solve the problem of fault feature extraction and fault type recognition were studied. The original signal was transformed into a set of sub-band wavelet coefficients using the TQWT. The effectiveness of extracting sparse wavelet coefficients using an iterative threshold shrinkage algorithm and the sensitivity of spectral kurtosis to fault impact signals were studied. By calculating the spectral kurtosis of each sub-band signal component and selecting the sub-band wavelet coefficient that contains obvious fault information, a fault feature extraction method for the sparse fault signal component was established. Using the sparse representation classification model of extracted fault signals, the method of rolling bearing fault-type recognition based on sparse representation was realized. Experimental results indicate that the proposed fault feature extraction method has a significant effect in eliminating interference components in the Case Western Reserve University dataset. The average diagnostic accuracy for the four types of data is 99.83%. The average diagnostic accuracy for the 10 types of data is 97.73%. Compared with the TQWT and iterative threshold shrinkage algorithm for fault feature extraction, the fault diagnosis accuracy of the proposed method improves by 11.60%, and the running time reduces by 8%. For the vibration dataset collected by the QPZZ-Ⅱ rotating machinery platform, the average diagnostic accuracy of the proposed method for the four types of data is 100%. Compared with the traditional wavelet denoising method, the accuracy of the proposed method improves by 35.67%, and the running time reduces by 7.25%. Therefore, the proposed method can effectively solve the problem of rolling-bearing fault diagnosis. 7 tabs, 7 figs, 30 refs.
-
表 1 第1组试验数据类型
Table 1. First group experimental data types
故障类型 故障尺寸/mm 电机转速/(r·min-1) 标签 无故障 0 1 797 1 滚动体故障 0.177 8 2 内圈故障 3 外圈故障 4 表 2 第2组试验数据类型
Table 2. Second group experimental data types
故障类型 故障尺寸/mm 电机转速/(r·min-1) 标签 无故障 0 1 797 1 内圈故障 0.177 8 1 797 2 内圈故障 1 772 3 内圈故障 1 750 4 外圈故障 1 797 5 外圈故障 1 772 6 外圈故障 1 750 7 滚动体故障 1 797 8 滚动体故障 1 772 9 滚动体故障 1 750 10 表 3 故障信号经TQWT后各子带谱峭度
Table 3. Kurtosis value of each subband of fault vibration signal after TQWT transformation
子带 1 2 3 4 5 6 7 8 9 10 11 正常信号谱峭度 8.41 22.34 18.77 3.65 6.77 64.74 5.92 2.98 4.33 14.52 3.88 滚动体故障信号谱峭度 12.24 3.31 7.89 48.06 26.97 8.67 9.69 22.89 23.37 8.75 5.06 内圈故障信号普峭度 20.71 7.95 10.05 19.06 30.02 13.02 6.85 12.95 58.92 6.49 5.10 外圈故障信号普峭度 38.48 7.07 4.28 10.13 58.20 13.26 12.40 21.65 30.39 12.72 6.61 表 4 故障识别结果
Table 4. Fault identification results
试验次数 1 2 3 4 5 6 7 8 9 10 10次试验平均值 平均分类准确率/% 100.00 100.00 99.72 99.44 100.00 99.72 99.72 100.00 100.00 99.72 99.83 正常数据分类准确率/% 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 滚动体分类准确率/% 100.00 100.00 98.89 97.78 100.00 98.89 98.89 100.00 100.00 98.89 99.33 内圈分类准确率/% 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 外圈分类准确率/% 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 表 5 五种方法故障识别结果对比
Table 5. Comparison of fault identification results for five kinds of methods
方法 准确率/% 运行时间/s 准确率标准偏差/% 1 95.30 0.81 1.02 2 95.75 0.75 1.36 3 97.47 0.70 0.84 4 98.28 0.62 0.79 5 99.83 0.68 0.18 表 6 三种方法故障识别结果对比
Table 6. Comparison of fault identification results for three kinds of methods
方法 准确率/% 运行时间/s 准确率标准偏差/% 1 86.13 4.03 1.48 4 86.38 3.86 0.77 5 97.73 3.68 0.73 表 7 两种方法故障识别结果对比
Table 7. Comparison of fault identification results for two kinds of methods
方法 准确率/% 运行时间/s 准确率标准偏差/% 4 64.33 0.69 2.57 5 100.00 0.64 0.00 -
[1] NANDI S, TOLIYAT H A, LI X. Condition monitoring and fault diagnosis of electrical motors—a review[J]. IEEE Transactions on Energy Conversion, 2005, 20(4): 719-729. doi: 10.1109/TEC.2005.847955 [2] 王国彪, 何正嘉, 陈雪峰, 等. 机械故障诊断基础研究"何去何从"[J]. 机械工程学报, 2013, 49(1): 63-72. https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201301010.htmWANG Guo-biao, HE Zheng-jia, CHEN Xue-feng, et al. Basic research on machinery fault diagnosis—what is the prescription[J]. Journal of Mechanical Engineering, 2013, 49(1): 63-72. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201301010.htm [3] 沈长青, 王旭, 王冬, 等. 基于多尺度卷积类内迁移学习的列车轴承故障诊断[J]. 交通运输工程学报, 2020, 20(5): 151-164. doi: 10.19818/j.cnki.1671-1637.2020.05.012SHEN Chang-qing, WANG Xu, WANG Dong, et al. Multi-scale convolution intra-class transfer learning for train bearing fault diagnosis[J]. Journal of Traffic and Transportation Engineering, 2020, 20(5): 151-164. (in Chinese) doi: 10.19818/j.cnki.1671-1637.2020.05.012 [4] 马小骏, 任淑红, 左洪福, 等. 基于LS-SVM算法和性能可靠性的航空发动机在翼寿命预测方法[J]. 交通运输工程学报, 2015, 15(3): 92-100. doi: 10.3969/j.issn.1671-1637.2015.03.013MA Xiao-jun, REN Shu-hong, ZUO Hong-fu, et al. Prediction method of aero-engine life on wing based on LS-SVM algorithm and performance reliability[J]. Journal of Traffic and Transportation Engineering, 2015, 15(3): 92-100. (in Chinese) doi: 10.3969/j.issn.1671-1637.2015.03.013 [5] KLEIN R, INGMAN D, BRAUN S. Non-stationary signals: phase-energy approach—theory and simulations[J]. Mechanical Systems and Signal Processing, 2001, 15(6): 1061-1089. doi: 10.1006/mssp.2001.1398 [6] FENG Zhi-peng, LIANG Ming, CHU Fu-lei. Recent advances in time-frequency analysis methods for machinery fault diagnosis: a review with application examples[J]. Mechanical Systems and Signal Processing, 2013, 38(1): 165-205. doi: 10.1016/j.ymssp.2013.01.017 [7] ANTONI J. Cyclostationarity by examples[J]. Mechanical Systems and Signal Processing, 2009, 23(4): 987-1036. doi: 10.1016/j.ymssp.2008.10.010 [8] LEI Ya-guo, HE Zheng-jia, ZI Yan-yang. Application of the EEMD method to rotor fault diagnosis of rotating machinery[J]. Mechanical Systems and Signal Processing, 2009, 23(4): 1327-1338. doi: 10.1016/j.ymssp.2008.11.005 [9] 李奕璠, 刘建新, 林建辉, 等. 基于自适应多尺度形态学分析的车轮扁疤故障诊断方法[J]. 交通运输工程学报, 2015, 15(1): 58-65. doi: 10.3969/j.issn.1671-1637.2015.01.008LI Yi-fan, LIU Jian-xin, LIN Jian-hui, et al. Fault diagnosis method of railway vehicle with wheel flat based on self-adaptive multi-scale morphology analysis[J]. Journal of Traffic and Transportation Engineering, 2015, 15(1): 58-65. (in Chinese) doi: 10.3969/j.issn.1671-1637.2015.01.008 [10] YAN R Q, GAO R X, CHEN X F. Wavelets for fault diagnosis of rotary machines: a review with applications[J]. Signal Processing, 2014, 96: 1-15. doi: 10.1016/j.sigpro.2013.04.015 [11] 张志禹, 吕延军, 张九龙, 等. 航空发动机转子碰摩故障瞬时频率诊断方法[J]. 交通运输工程学报, 2007, 7(4): 21-23. doi: 10.3321/j.issn:1671-1637.2007.04.005ZHANG Zhi-yu, LYU Yan-jun, ZHANG Jiu-long, et al. Diagnosis method of instantaneous frequency for rotor impact-rub fault of aeroengine[J]. Journal of Traffic and Transportation Engineering, 2007, 7(4): 21-23. (in Chinese) doi: 10.3321/j.issn:1671-1637.2007.04.005 [12] SELESNICK I W. Wavelet transform with tunable Q-factor[J]. IEEE Transactions on Signal Processing, 2011, 59(8): 3560-3575. doi: 10.1109/TSP.2011.2143711 [13] LIU Ruo-nan, YANG Bo-yuan, ZIO E, et al. Artificial intelligence for fault diagnosis of rotating machinery: a review[J]. Mechanical Systems and Signal Processing, 2018, 108: 33-47. doi: 10.1016/j.ymssp.2018.02.016 [14] ZHANG Shen, ZHANG Shi-bo, WANG Bin-guan, et al. Deep learning algorithms for bearing fault diagnostics—a comprehensive review[J]. IEEE Access, 2020, 8: 29857-29881. doi: 10.1109/ACCESS.2020.2972859 [15] 高倩, 陈晓英, 孙丽颖. 基于稀疏表示的TQWT在低频振荡信号去噪中应用[J]. 电力系统保护与控制, 2016, 44(13): 55-60. doi: 10.7667/PSPC151358GAO Qian, CHEN Xiao-ying, SUN Li-ying. Low frequency oscillating signals denoising based on TQWT via sparse representation[J]. Power System Protection and Control, 2016, 44(13): 55-60. (in Chinese) doi: 10.7667/PSPC151358 [16] 王霄, 谢平, 郭源耕, 等. 基于多字典-共振稀疏分解的脉冲故障特征提取[J]. 中国机械工程, 2019, 30(20): 2456-2462, 2472. doi: 10.3969/j.issn.1004-132X.2019.20.008WANG Xiao, XIE Ping, GUO Yuan-geng, et al. Impulse fault signature extraction based on multi-dictionary resonance-based sparse signal decomposition[J]. China Mechanical Engineering, 2019, 30(20): 2456-2462, 2472. (in Chinese) doi: 10.3969/j.issn.1004-132X.2019.20.008 [17] CAI Gai-gai, CHEN Xue-feng, HE Zheng-jia. Sparsity-enabled signal decomposition using tunable Q-factor wavelet transform for fault feature extraction of gearbox[J]. Mechanical Systems and Signal Processing, 2013, 41(1/2): 34-53. [18] ZHAO Zhi-bin, WANG Shi-bin, AN Bo-tao, et al. Hierarchical hyper- laplacian prior for weak fault feature enhancement[J]. ISA Transactions, 2020, 96: 429-443. doi: 10.1016/j.isatra.2019.06.007 [19] 赵见龙, 张永超, 王立夫, 等. 基于共振稀疏分解与谱峭度的滚动轴承故障诊断[J]. 组合机床与自动化加工技术, 2019(4): 111-115. https://www.cnki.com.cn/Article/CJFDTOTAL-ZHJC201904027.htmZHAO Jian-long, ZHANG Yong-chao, WANG Li-fu, et al. Rolling bearing fault diagnosis based on resonance sparse decomposition and spectral kurtosis[J]. Modular Machine Tool and Automatic Manufacturing Technique, 2019(4): 111-115. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZHJC201904027.htm [20] 孔运, 王天杨, 褚福磊. 自适应TQWT滤波器算法及其在冲击特征提取中的应用[J]. 振动与冲击, 2019, 38(11): 9-16, 23. https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201911003.htmKONG Yun, WANG Tian-yang, CHU Fu-lei. Adaptive TQWT filter algorithm and its application in impact feature extraction[J]. Journal of Vibration and Shock, 2019, 38 (11): 9-16, 23. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201911003.htm [21] LI Jun-lin, WANG Hua-qing, SONG Liu-yang, et al. A novel feature extraction method for roller bearing using sparse decomposition based on self-adaptive complete dictionary[J]. Measurement, 2019, 148: 106934. doi: 10.1016/j.measurement.2019.106934 [22] FENG Z P, ZHOU Y K, ZUO M J, et al. Atomic decomposition and sparse representation for complex signal analysis in machinery fault diagnosis: a review with examples[J]. Measurement, 2017, 103: 106-132. doi: 10.1016/j.measurement.2017.02.031 [23] 王宏超, 陈进, 董广明, 等. 可调品质因子小波变换在转子早期碰摩故障诊断中应用[J]. 振动与冲击, 2014, 33(10): 77-80. https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201410015.htmWANG Hong-chao, CHEN Jin, DONG Guang-ming, et al. Early rub-impact diagnosis of rotors based on tunable Q-factor wavelet transformation[J]. Journal of Vibration and Shock, 2014, 33(10): 77-80. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201410015.htm [24] BECK A, TEBOULLE M. A fast iterative shrinkage-thresholding algorithm for linear inverse problems[J]. SIAM Journal on Imaging Sciences, 2009, 2(1): 183-202. doi: 10.1137/080716542 [25] 何正嘉, 訾艳阳, 陈雪峰, 等. 内积变换原理与机械故障诊断[J]. 振动工程学报, 2007, 20(5): 528-533. doi: 10.3969/j.issn.1004-4523.2007.05.019HE Zheng-jia, ZI Yan-yang, CHEN Xue-feng, et al. Transform principle of inner product for fault diagnosis[J]. Journal of Vibration Engineering, 2007, 20(5): 528-533. (in Chinese) doi: 10.3969/j.issn.1004-4523.2007.05.019 [26] 苏文胜, 王奉涛, 张志新, 等. EMD降噪和谱峭度法在滚动轴承早期故障诊断中的应用[J]. 振动与冲击, 2010, 29(3): 18-21. doi: 10.3969/j.issn.1000-3835.2010.03.005SU Wen-sheng, WANG Feng-tao, ZHANG Zhi-xin, et al. Application of EMD denoising and spectral kurtosis in early fault diagnosis of rolling element bearings[J]. Journal of Vibration and Shock, 2010, 29(3): 18-21. (in Chinese) doi: 10.3969/j.issn.1000-3835.2010.03.005 [27] ANTONI J. The spectral kurtosis: a useful tool for characterising non-stationary signals[J]. Mechanical Systems and Signal Processing, 2006, 20(2): 282-307. doi: 10.1016/j.ymssp.2004.09.001 [28] WRIGHT J, YANG A Y, GANESH A, et al. Robust face recognition via sparse representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(2): 210-227. doi: 10.1109/TPAMI.2008.79 [29] LI Yong-bo, WANG Xian-zhi, SI Shu-bin, et al. Entropy based fault classification using the case western reserve university data: a benchmark study[J]. IEEE Transactions on Reliability, 2020, 69(2): 754-767. doi: 10.1109/TR.2019.2896240 [30] SU W S, WANG F T, ZHU H, et al. Rolling element bearing faults diagnosis based on optimal Morlet wavelet filter and autocorrelation enhancement[J]. Mechanical Systems and Signal Processing, 2010, 24(5): 1458-1472. doi: 10.1016/j.ymssp.2009.11.011