Adaptive dynamic mode decomposition and GA-SVM with application to fault classification of planetary bearing
-
摘要: 行星齿轮箱在运行过程中由于齿轮间的相互作用会产生强噪声，导致行星轴承的故障特征被完全淹没在背景噪声中并难以提取，从而使得行星轴承故障分类的准确率较低。本文提出一种自适应动模式分解(ADMD)和遗传算法优化支持向量机(GA-SVM)的行星轴承故障分类方法。首先，针对传统动模式分解(DMD)中截断秩无法准确选取的问题，定义了一种新的适应度函数，并采用改进的蚱蜢优化算法(IGOA)自适应选取最优截断秩，进而实现对原始振动信号的降噪处理。然后对处理后的信号计算其归一化后的复合精细多尺度离散熵(IRCMDE)并构成特征矩阵。最后采用遗传算法优化支持向量机，构建GA-SVM分类模型，并将其应用到行星轴承故障诊断中。利用行星齿轮箱中行星轴承故障数据验证了此方法的有效性和实用性，最终分类结果为96.43%，表明了该方法可以准确识别出行星轴承的故障类型。Abstract: Recently, planetary gearboxes have been widely used in helicopters, heavy trucks, ships, and other large and complex mechanical equipment because of their smooth transmission characteristics, small volume, and large reduction ratio. The planetary bearing, which plays a supporting role in the planetary gearbox, usually works in a worse environment but suffers from low speed and heavy load for a long time. Additionally, because of the strong noise generated by the interaction between gears during the operation of the planetary gearbox, the fault characteristics of planetary bearings are completely submerged in the background noise and are difficult to extract, which complicates classifying planetary-bearing faults accurately. Therefore, to effectively remove noise information from planetary-bearing signals, accurately extract fault information, and classify the fault types of planetary bearings, an adaptive dynamic mode decomposition (ADMD) and genetic algorithm and support vector machine (GA-SVM) with application to the fault classification of planetary bearing is proposed in this paper. The hard threshold selection of the traditional truncated rank cannot effectively process the time-domain vibration signals using the dynamic mode decomposition (DMD) method. Hence, this paper proposes improved grasshopper optimization algorithm (IGOA) to optimize the grasshopper optimization algorithm (GOA) by using dynamic weight and avoid the linear gradient mechanism, which cannot fully use the entire iterative process. Furthermore, IGOA can perform a global search to achieve the adaptive optimal parameter selection of the truncated rank. Besides, a new fitness function is defined that can effectively process the original time-domain signals. The traditional refined composite multiscale discrete entropy (RCMDE) is relatively dispersed, and it cannot characterize the features hidden in the signal better. Therefore, we normalize the RCMDE, forming the improved refined composite multiscale discrete entropy (IRCMDE). Then, the IRCMDE is calculated for the denoised signal, and a feature matrix is constructed to better mine the hidden features in the signal. Finally, GA is used to optimize the key parameters C and g of the SVM. The GA-SVM classification model is also constructed and applied to the bearing fault classification of the planetary gearbox, which can avoid the overfitting phenomenon in the training process and provide better generalization performance. Taking the planetary-bearing fault data in the planetary gearbox of Nanchang Hangkong University as the research object, the validity and practicability of the proposed method are verified, and the final classification result of the inner ring fault, outer ring fault, rolling body fault, and normal condition is 96.43%. In addition, this method can more accurately identify the fault types of planetary bearings and has better generalization ability than the empirical mode decomposition (EMD) signal processing method and the convolutional neural network (CNN) classification method.
-
图 4 行星齿轮箱传感器布置图. (a) 行星齿轮箱; (b) 传感器位置
Figure 4. Planetary gear box sensor layout: (a) planetary gear box; (b) sensor location
图 5 不同故障模式下振动信号时域图. (a) 内圈信号; (b) 外圈信号; (c) 滚动体信号; (d) 正常信号
Figure 5. Time-domain diagrams of vibration signals under different fault modes: (a) inner ring signal; (b) outer ring signal; (c) rolling body signal; (d) normal signal
图 7 不同方法分类结果. (a) ADMD+GA-SVM法; (b) ADMD+CNN法; (c) EMD+GA-SVM法
Figure 7. Classification results by different methods: (a) ADMD+GA-SVM method; (b) ADMD+CNN method; (c) EMD+GA-SVM method
表 1 行星齿轮箱的主要参数
Table 1. Main parameters of planetary gear box
Number of teeth
of sun wheelNumber of planetary
gear teethRing gear 28 36 (3) 100 
下载: 导出CSV
表 2 不同方法对行星齿轮箱行星轴承分类结果
Table 2. Classification results of planetary bearings in planetary gearboxes by different methods
Methods Accuracy /% ADMD+GA-SVM 96.43 ADMD+CNN 78.57 EMD+GA-SVM 91.07
下载: 导出CSV
-
参考文献
[1] Lei Y G. Adaptive ensemble empirical mode decomposition and its application to fault detection of planetary gearboxes. J Mech Eng, 2014, 50(3): 64 doi: 10.3901/JME.2014.03.064 [2] Wang T Y, Chu F L, Feng Z P. Meshing frequency modulation (MFM) index-based kurtogram for planet bearing fault detection. J Sound Vib, 2018, 432: 437 doi: 10.1016/j.jsv.2018.06.051 [3] Wang P, Li T Y, Gao X J, et al. Bearing fault signal denoising method of hierarchical adaptive wavelet threshold function. J Vib Eng, 2019, 32(3): 548 doi: 10.16385/j.cnki.issn.1004-4523.2019.03.021王普, 李天垚, 高学金, 等. 分层自适应小波阈值轴承故障信号降噪方法. 振动工程学报, 2019, 32(3):548 doi: 10.16385/j.cnki.issn.1004-4523.2019.03.021 [4] Zhao X M, Patel T H, Zuo M J. Multivariate EMD and full spectrum based condition monitoring for rotating machinery. Mech Syst Signal Process, 2012, 27: 712 doi: 10.1016/j.ymssp.2011.08.001 [5] Wang Z Y, Yao L G, Qi X L, et al. Fault diagnosis of planetary gearbox based on parameter optimized VMD and multi-domain manifold learning. J Vib Shock, 2021, 40(1): 110 doi: 10.13465/j.cnki.jvs.2021.01.015王振亚, 姚立纲, 戚晓利, 等. 参数优化变分模态分解与多域流形学习的行星齿轮箱故障诊断. 振动与冲击, 2021, 40(1):110 doi: 10.13465/j.cnki.jvs.2021.01.015 [6] Zheng X W, Tang Y Y, Zhou J T. A framework of adaptive multiscale wavelet decomposition for signals on undirected graphs. IEEE Trans Signal Process, 2019, 67(7): 1696 doi: 10.1109/TSP.2019.2896246 [7] Lv Y, Yuan R, Wang T, et al. Health degradation monitoring and early fault diagnosis of a rolling bearing based on CEEMDAN and improved MMSE. Materials, 2018, 11(6): 1009 doi: 10.3390/ma11061009 [8] Zheng Y, Yue J H, Jiao J, et al. Fault feature extraction method of rolling bearing based on parameter optimized VMD. J Vib Shock, 2021, 40(1): 86 doi: 10.13465/j.cnki.jvs.2021.01.012郑义, 岳建海, 焦静, 等. 基于参数优化变分模态分解的滚动轴承故障特征提取方法. 振动与冲击, 2021, 40(1):86 doi: 10.13465/j.cnki.jvs.2021.01.012 [9] Dang Z, Lv Y, Li Y R, et al. A fault diagnosis method for one-dimensional vibration signal based on multiresolution tlsDMD and approximate entropy. Shock Vib, 2019, 2019: 1 [10] Rowley C W, Mezić I, Bagheri S, et al. Spectral analysis of nonlinear flows. J Fluid Mech, 2009, 641: 115 doi: 10.1017/S0022112009992059 [11] Taira K, Brunton S L, Dawson S T M, et al. Modal analysis of fluid flows: An overview. AIAA J, 2017, 55(12): 4013 doi: 10.2514/1.J056060 [12] Schmid P J, Li L, Juniper M P, et al. Applications of the dynamic mode decomposition. Theor Comput Fluid Dyn, 2011, 25(1): 249 [13] Dou D, Ye X D, Zhang G H. Entropy sequences and maximal entropy sets. Nonlinearity, 2006, 19(1): 53 doi: 10.1088/0951-7715/19/1/004 [14] Pincus S M. Approximate entropy as a measure of system complexity. Proc Natl Acad Sci, 1991, 88(6): 2297 doi: 10.1073/pnas.88.6.2297 [15] Shi Y, Lin J H, Zhuang Z, et al. Fault diagnosis for pantograph cracks based on time-frequency decomposition and sample entropy of vibration signals. J Vib Shock, 2019, 38(8): 180 doi: 10.13465/j.cnki.jvs.2019.08.027施莹, 林建辉, 庄哲, 等. 基于振动信号时频分解-样本熵的受电弓裂纹故障诊断. 振动与冲击, 2019, 38(8):180 doi: 10.13465/j.cnki.jvs.2019.08.027 [16] Dang Z, Lv Y, Li Y R, et al. Optimized dynamic mode decomposition via non-convex regularization and multiscale permutation entropy. Entropy, 2018, 20(3): 152 doi: 10.3390/e20030152 [17] Lin L, Zhao D Y. Application of approximate entropy in acoustic emission signal processing. J Vib Shock, 2008, 27(2): 99 doi: 10.3969/j.issn.1000-3835.2008.02.023林丽, 赵德有. 近似熵在声发射信号处理中的应用. 振动与冲击, 2008, 27(2):99 doi: 10.3969/j.issn.1000-3835.2008.02.023 [18] Jiang Y, Mao D, Xu Y S. A fast algorithm for computing sample entropy. Adv Adapt Data Anal, 2011, 3(1−2): 167 doi: 10.1142/S1793536911000775 [19] Azami H, Escudero J. Amplitude-aware permutation entropy: Illustration in spike detection and signal segmentation. Comput Methods Programs Biomed, 2016, 128: 40 doi: 10.1016/j.cmpb.2016.02.008 [20] Rostaghi M, Azami H. Dispersion entropy: A measure for time-series analysis. IEEE Signal Process Lett, 2016, 23(5): 610 doi: 10.1109/LSP.2016.2542881 [21] Azami H, Rostaghi M, Abásolo D, et al. Refined composite multiscale dispersion entropy and its application to biomedical signals. IEEE Trans Biomed Eng, 2017, 64(12): 2872 doi: 10.1109/TBME.2017.2679136 [22] Achlerkar P D, Samantaray S R, Sabarimalai Manikandan M. Variational mode decomposition and decision tree based detection and classification of power quality disturbances in grid-connected distributed generation system. IEEE Trans Smart Grid, 2018, 9(4): 3122 doi: 10.1109/TSG.2016.2626469 [23] Zhang S Q, Huang W J, Hu Y T, et al. Bearing fault diagnosis method based on EEMD approximate entropy and hybrid PSO–BP algorithm. China Mech Eng, 2016, 27(22): 3048 doi: 10.3969/j.issn.1004-132X.2016.22.012张淑清, 黄文静, 胡永涛, 等. 基于总体平均经验模式分解近似熵和混合PSO-BP算法的轴承故障诊断方法. 中国机械工程, 2016, 27(22):3048 doi: 10.3969/j.issn.1004-132X.2016.22.012 [24] Abdoos A A, Mianaei P K, Ghadikolaei M R. Combined VMD-SVM based feature selection method for classification of power quality events. Appl Soft Comput, 2016, 38(C): 637 [25] Saremi S, Mirjalili S, Lewis A. Grasshopper optimisation algorithm: Theory and application. Adv Eng Softw, 2017, 105: 30 doi: 10.1016/j.advengsoft.2017.01.004 [26] Meraihi Y, Benmessaoud Gabis A, Mirjalili S, et al. Grasshopper optimization algorithm: Theory, variants, and applications. IEEE Access, 2021, 9: 50001 doi: 10.1109/ACCESS.2021.3067597 [27] Zhao R, Ni H, Feng H W, et al. A dynamic weight grasshopper optimization algorithm with random jumping // Advances in Intelligent Systems and Computing. Singapore: Springer Singapore, 2019: 401 [28] Yu J B, Hu T Z, Liu H Q. A new morphological filter for fault feature extraction of vibration signals. IEEE Access, 2019, 7: 53743 doi: 10.1109/ACCESS.2019.2912898 [29] Ma Y B, Lv Y, Yuan R, et al. Matching synchroextracting transform for mechanical fault diagnosis under variable-speed conditions. IEEE Trans Instrum Meas, 2022, 71: 1 [30] Wang D. Some further thoughts about spectral kurtosis, spectral L2/L1 norm, spectral smoothness index and spectral Gini index for characterizing repetitive transients. Mech Syst Signal Process, 2018, 108: 360 doi: 10.1016/j.ymssp.2018.02.034 [31] Ren Y, Li W, Zhang B, et al. Fault diagnosis of rolling bearings based on improved kurtogram in varying speed conditions. Appl Sci, 2019, 9(6): 1157 doi: 10.3390/app9061157 [32] Zhang S, Zhang Y X, Zhu J P. Rolling element-bearing feature extraction based on combined wavelets and quantum-behaved particle swarm optimization. J Mech Sci Technol, 2015, 29(2): 605 doi: 10.1007/s12206-015-0120-3 [33] Tang X B, Wang Z Q, Zhong L X. Microblog topic tracking model based on wikipedia semantic extension. Inf Sci, 2017, 35(2): 80 doi: 10.13833/j.cnki.is.2017.02.014唐晓波, 王中勤, 钟林霞. 基于维基语义扩展的微博话题追踪模型研究. 情报科学, 2017, 35(2):80 doi: 10.13833/j.cnki.is.2017.02.014 [34] Zhang Y, Wu L H, Fan X M, et al. Identification and classification of power quality disturbances based on modified S transform and GA-SVM. Guangdong Electr Power, 2021, 34(5): 99张殷, 武利会, 范心明, 等. 基于改进S变换和GA-SVM的电能质量扰动识别与分类. 广东电力, 2021, 34(5):99 [35] Xu Y, Song W X, Feng J Y, et al. Minimum pulse width modulation of current harmonic for three-level inverter. J Shanghai Univ Nat Sci Ed, 2017, 23(5): 690徐渊, 宋文祥, 冯九一, 等. 三电平逆变器电流谐波最小PWM方法. 上海大学学报(自然科学版), 2017, 23(5):690 -
点击查看大图
图(7) / 表(2)
计量
- 文章访问数: 495
- HTML全文浏览量: 147
- PDF下载量: 96
- 被引次数: 0
