Performance Degradation Assessment of Gears based on AR Model and Dictionary Learning
-
摘要: 齿轮性能退化评估是预诊断的提前和基础,针对概率相似度量评估方法存在模型复杂,容易过早饱和等现象,提出一种基于AR (Autoregressive model)模型和字典学习的齿轮性能退化评估的重构模型方法,其中AR模型用于提取齿轮振动信号的状态特征,字典学习通过正常状态下构建的字典模型(Dictionary learning, DL)对测试样本进行AR模型系数重构。首先提取正常运行状态下振动信号的AR模型系数构建过完备字典模型,然后将待测信号的AR系数作为特征向量输入字典模型中得到重构后的AR模型系数。最后由原始AR系数和重构AR系数分别构造自回归模型,并各自完成对待测信号的时序建模,将两自回归模型所得残差序列的均方根误差作为性能劣化程度指标。全寿命疲劳实验数据分析结果表明,与传统时域指标相比该方法对早期故障更敏感且具有与齿轮故障发展趋势一致性更好等优点。Abstract: The performance degradation assessment (PDA) is the advance and foundation of gear fault pre-diagnosis. Aiming at the problems that similarity-based methods are of complex models and time consuming, a reconstruction model of gear PDA based on Autoregressive Model (AR) and Dictionary Learning (DL) is proposed. The coefficients of AR model serve as feature vectors to depict gear performance states and DL model is used to reconstruct AR coefficients. Firstly, the over-complete dictionary model are constructed with the AR coefficients of vibration signals under normal operation, and then the AR coefficients of the signals under consideration are input into the DL model as feature vectors to obtain the reconstructed feature vectors. By inputting the signal at hand into the two autoregressive models composed of original AR coefficients and reconstructed AR coefficients respectively, the corresponding residual sequences are obtained. Finally, Mean Squared Error (MSE) is introduced to evaluate the two residual sequences as performance degradation index. Run-to-failure data set from gear are processed to demonstrate the advantages of the method in terms of the trendability, consistency and sensitivity of early failure.
-
Key words:
- ARmodel /
- dictionary learning /
- gear /
- performance degradation assessment /
- restructuring model
-
表 1 圆柱齿轮齿轮箱参数
名称 主动轮(测试齿轮) 从动轮 转速/(r·min−1) 1000 1000 载荷/(daN·m) 200 200 齿数 20 21 转频/Hz 16.67 15.86 齿宽/m 0.015 0.03 模数/m 0.01 0.01 压力角/(°) 20 20 表 2 齿轮箱疲劳试验每日停机观测结果
观测日期/d 观测结果 1 无故障 2 无故障 3 无故障 4 无故障 5 无故障 6 第二齿产生剥落 7 第二齿剥落程度无恶化 8 第二齿无恶化;第十六齿产生早期剥落 9 第十六齿剥落继续增大 10 同上 11 同上 12 第十六齿的剥落面积覆盖到整个齿宽 -
[1] LIU J B, DJURDJANOVIC D, NI J, et al. Similarity based method for manufacturing process performance prediction and diagnosis[J]. Computers in Industry, 2007, 58(6): 558-566 doi: 10.1016/j.compind.2006.12.004 [2] 周东华, 魏慕恒, 司小胜. 工业过程异常检测、寿命预测与维修决策的研究进展[J]. 自动化学报, 2013, 39(6): 711-722ZHOU D H, WEI M H, SI X S. A survey on anomaly detection, life prediction and maintenance decision for industrial processes[J]. Acta Automatica Sinica, 2013, 39(6): 711-722 (in Chinese) [3] 张龙, 成俊良, 杨世锡, 等. 基于时序模型和自联想神经网络的齿轮故障程度评估[J]. 振动与冲击, 2019, 38(2): 18-24ZHANG L, CHENG J L, YANG S X, et al. Fault severity assessment for gears based on AR model and auto-associative neural network[J]. Journal of Vibration and Shock, 2019, 38(2): 18-24 (in Chinese) [4] HONARVAR F, MARTIN H R. New statistical moments for diagnostics of rolling element bearings[J]. Journal of Manufacturing Science and Engineering, 1997, 119(3): 425-432 doi: 10.1115/1.2831123 [5] 罗毅, 甄立敬. 基于小波包与倒频谱分析的风电机组齿轮箱齿轮裂纹诊断方法[J]. 振动与冲击, 2015, 34(3): 210-214LUO Y, ZHEN L J. Diagnosis method of turbine gearbox gearcrack based on wavelet packet and cepstrum analysis[J]. Journal of Vibration and Shock, 2015, 34(3): 210-214 (in Chinese) [6] TANG B P, LIU W Y, SONG T. Wind turbine fault diagnosis based on Morlet wavelet transformation and Wigner-Ville distribution[J]. Renewable Energy, 2010, 35(12): 2862-2866 doi: 10.1016/j.renene.2010.05.012 [7] 马伦, 康建设, 孟妍, 等. 基于Morlet小波变换的滚动轴承早期故障特征提取研究[J]. 仪器仪表学报, 2013, 34(4): 920-926 doi: 10.3969/j.issn.0254-3087.2013.04.031MA L, KANG J S, MENG Y, et al. Research on feature extraction of rolling bearing incipient fault based on Morlet wavelet transform[J]. Chinese Journal of Scientific Instrument, 2013, 34(4): 920-926 (in Chinese) doi: 10.3969/j.issn.0254-3087.2013.04.031 [8] 于德介, 程军圣, 杨宇. 基于EMD和AR模型的滚动轴承故障诊断方法[J]. 振动工程学报, 2004, 17(3): 332-335 doi: 10.3969/j.issn.1004-4523.2004.03.016YU D J, CHENG J S, YANG Y. A fault diagnosis approach for roller bearings based on EMD method and AR model[J]. Journal of Vibration Engineering, 2004, 17(3): 332-335 (in Chinese) doi: 10.3969/j.issn.1004-4523.2004.03.016 [9] 孙国富, 徐玉秀. 应用AR模型的多参数与多测点信息融合的故障分类[J]. 机械科学与技术, 2017, 36(6): 925-932SUN G F, XU Y X. Fault classification based on multi-parameter and multi-point information fusion of AR model[J]. Mechanical Science and Technology for Aerospace Engineering, 2017, 36(6): 925-932 (in Chinese) [10] LI Z X, YAN X P, YUAN C Q, et al. Virtual prototype and experimental research on gear multi-fault diagnosis using wavelet-autoregressive model and principal component analysis method[J]. Mechanical Systems and Signal Processing, 2011, 25(7): 2589-2607 doi: 10.1016/j.ymssp.2011.02.017 [11] 冯辅周, 司爱威, 江鹏程. 小波相关排列熵和HMM在故障预测中的应用[J]. 振动工程学报, 2013, 26(2): 269-276 doi: 10.3969/j.issn.1004-4523.2013.02.017FENG F Z, SI A W, JIANG P C. Application of wavelet correlation permutation entropy and Hidden Markov Model to fault prognostic[J]. Journal of Vibration Engineering, 2013, 26(2): 269-276 (in Chinese) doi: 10.3969/j.issn.1004-4523.2013.02.017 [12] HEYNS T, HEYNS P S, DE VILLIERS J P. Combining synchronous averaging with a Gaussian mixture model novelty detection scheme for vibration-based condition monitoring of a gearbox[J]. Mechanical Systems and Signal Processing, 2012, 32: 200-215 doi: 10.1016/j.ymssp.2012.05.008 [13] ZHAN Y M, MAKIS V. A robust diagnostic model for gearboxes subject to vibration monitoring[J]. Journal of Sound and Vibration, 2006, 290(3-5): 928-955 doi: 10.1016/j.jsv.2005.04.018 [14] ZHAN Y M, MECHEFSKE C K. Robust detection of gearbox deterioration using compromised autoregressive modeling and Kolmogorov–Smirnov test statistic Part II: Experiment and application[J]. Mechanical Systems and Signal Processing, 2007, 21(5): 1983-2011 doi: 10.1016/j.ymssp.2006.11.006 [15] HUBEL D H, WIESEL T N. Receptive fields of single neurones in the cat's striate cortex[J]. The Journal of Physiology, 1959, 148(3): 574-591 doi: 10.1113/jphysiol.1959.sp006308 [16] 周海韬. 基于字典学习理论的机械故障诊断方法研究[D]. 上海: 上海交通大学, 2016ZHOU H T. Machinery fault diagnosis method based on dictionary learning theory[D]. Shanghai: Shanghai Jiao Tong University, 2016 (in Chinese) [17] 朱会杰, 王新晴, 芮挺, 等. 多尺度移不变稀疏编码及其在机械故障诊断中的应用[J]. 北京理工大学学报, 2016, 36(1): 19-24ZHU H J, WANG X Q, RUI T, et al. Multi scale shift invariant sparse coding for robust machinery diagnosis[J]. Transactions of Beijing Institute of Technology, 2016, 36(1): 19-24 (in Chinese) [18] 苗中华, 周广兴, 刘海宁, 等. 基于稀疏编码的振动信号特征提取算法与实验研究[J]. 振动与冲击, 2014, 33(15): 76-81+118MIAO Z H, ZHOU G X, LIU H N, et al. Tests and feature extraction algorithm of vibration signals based on sparse coding[J]. Journal of Vibration and Shock, 2014, 33(15): 76-81+118 (in Chinese) [19] 何翔, 高宏力, 郭亮, 等. 基于AR模型和谱熵的自适应小波包络检测[J]. 中国机械工程, 2017, 28(3): 348-352 doi: 10.3969/j.issn.1004-132X.2017.03.016HE X, GAO H L, GUO L, et al. Adaptive wavelet envelope detection based on AR model and spectral entropy[J]. China Mechanical Engineering, 2017, 28(3): 348-352 (in Chinese) doi: 10.3969/j.issn.1004-132X.2017.03.016 [20] 王聪. 基于稀疏表达的机械信号处理方法及其在滚动轴承故障诊新中的应用研究[D]. 合肥: 中国科学技术大学, 2017WANG C. Mechanical siganl processing based on sparse representation and its application in the diagnosis of rolling-element bearings[D]. Hefei: University of Science and Technology of China, 2017 (in Chinese) [21] DONOHO D L, HUO X. Uncertainty principles and ideal atomic decomposition[J]. IEEE Transactions on Information Theory, 2001, 47(7): 2845-2862 doi: 10.1109/18.959265 [22] ENGAN K, RAO B D, KREUTZ-DELGADO K. Frame design using FOCUSS with method of optimal directions (MOD)[C]// Proceedings of the 1999 Norwegian Signal Processing Symposium, 1999, 99: 65-69 [23] MALLAT S G, ZHANG Z F. Matching pursuits with time-frequency dictionaries[J]. IEEE Transactions on Signal Processing, 1993, 41(12): 3397-3415 doi: 10.1109/78.258082 [24] PATI Y C, REZAIIFAR R, KRISHNAPRASAD P S. Orthogonal matching pursuit: recursive function approximation with applications to wavelet decomposition[C]// Proceedings of 27th Asilomar Conference on Signals, Systems and Computers. Pacific Grove: IEEE, 1993: 40-44 [25] PAREY A, EL BADAOUI M, GUILLET F, et al. Dynamic modelling of spur gear pair and application of empirical mode decomposition-based statistical analysis for early detection of localized tooth defect[J]. Journal of Sound and Vibration, 2006, 294(3): 547-561 doi: 10.1016/j.jsv.2005.11.021 [26] 赵川. 特征降维与自适应特征提取方法及其在行星齿轮箱故障诊断中的应用研究[D]. 北京: 北京科技大学, 2018ZHAO C. Feature dimension reduction and adaptive extraction methods for planetary gearbox fault diagnosis[D]. Beijing: University of Science & Technology Beijing, 2018 (in Chinese) [27] 刘韬. 基于隐马尔可夫模型与信息融合的设备故障诊断与性能退化评估研究[D]. 上海: 上海交通大学, 2014LIU T. Study of hidden Markov model and information fusion in equipment fault diagnosis and performance degradation assessment[D]. Shanghai: Shanghai Jiao tong University, 2014 (in Chinese)