SVDD Fault Diagnosis Method of Planetary Gears based on Fitting Function Coefficient
-
摘要: 行星齿轮振动信号复杂多变,离散型故障信号需进行降噪、降维等复杂的处理过程,且信号处理过程中易造成信息缺失等。针对以上存在的问题,提出将函数型数据分析(Functional data analysis,FDA)与改进的支持向量数据描述(Support vector domain description,SVDD)相结合的故障诊断方法。根据齿轮不同故障类型建立不同的故障拟合基函数,将训练集数据与傅里叶基函数进行拟合,根据拟合得到的函数系数特征建立SVDD模型,并以ROC(Receiver operating characteristic)的评价函数为优化目标使用模拟退火算法对SVDD模型中的核参数σ和惩罚因子c进行优化;将不同的测试样本带入SVDD模型中,通过计算测试样本到超球体球心的相对距离来识别故障种类,进而完成行星齿轮的故障诊断。实验结果对比表明,本文中提出的方法能够解决离散型故障信号处理复杂、信息丢失等问题,准确地识别行星齿轮故障种类。Abstract: For the problems of complex and variable vibration signals of planetary gears, feature extraction distortion and small sample size, a new fault diagnosis method combining function data analysis with improved support vector domain description (SVDD) is proposed. According to different fault types of gears, different fault fitting basis functions are established, the training set data is fitted with the Fourier basis function, and the SVDD model is established according to the function coefficient features obtained by fitting, and in a ROC (receiver operating characteristic) of the evaluation function optimization objective simulated annealing algorithm kernel parameter model SVDD penalty factors σ/c and optimized; different test samples are taken into the SVDD model, and the fault type is determined by calculating the relative distance between the test sample and the center of the ball, thereby completing the planetary gear fault diagnosis. The comparison of experimental results shows that the proposed method can accurately identify the types of planetary gear faults.
-
表 1 行星齿轮参数
齿轮种类 太阳轮 行星轮(个数) 齿圈 齿数/个 27 40(3) 108 表 2 行星齿轮参数
啮合频率/Hz 旋转频率/Hz 故障特征频率/Hz 齿圈 太阳轮 行星架 太阳轮 行星轮 864 40 8 96 21.6 24 表 3 不同故障诊断模型的分类结果
状态 诊断模型 参数 正确识别样本累计总数 未正确识别的样本累计总数 测试精度 参数优化前 FDA+SVDD σ=64, c=0.1 56 32 63.75% 时域特征+SVDD σ=30, c=0.5 59 35 63.47% FDA+KNN k=11 42 18 58.57% 时域特征+KNN k=11 23 37 48.43% 参数优化后 FDA+SVDD σ=2.563, c=0.334 57 3 92.59% 时域特征+SVDD σ=2.632, c=0.325 54 10 82.16% FDA+KNN k=23 53 7 83.02% 时域特征+KNN k=23 51 9 78.37% -
[1] 王亚萍, 匡宇麒, 葛江华, 等.CEEMD和小波半软阈值相结合的滚动轴承降噪[J].振动、测试与诊断, 2018, 38(1):80-86 http://d.old.wanfangdata.com.cn/Periodical/zdcsyzd201801012Wang Y P, Kuang Y Q, Ge J H, et al. De-noising method for bearing vibration signal based on CEEMD and wavelet semi-soft threshold[J]. Journal of Vibration, Measurement & Diagnosis, 2018, 38(1):80-86(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdcsyzd201801012 [2] 雷亚国, 何正嘉, 林京, 等.行星齿轮箱故障诊断技术的研究进展[J].机械工程学报, 2011, 47(19):59-67 http://d.old.wanfangdata.com.cn/Periodical/jxgcxb201119008Lei Y G, He Z J, Lin J, et al. Research advances of fault diagnosis technique for planetary gearboxes[J]. Journal of Mechanical Engineering, 2011, 47(19):59-67(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jxgcxb201119008 [3] 冯志鹏, 赵镭镭, 褚福磊.行星齿轮箱齿轮局部故障振动频谱特征[J].中国电机工程学报, 2013, 33(5):119-127 http://d.old.wanfangdata.com.cn/Periodical/zgdjgcxb201305017Feng Z P, Zhao L L, Chu F L. Vibration spectral characteristics of localized gear fault of planetary gearboxes[J]. Proceedings of the CSEE, 2013, 33(5):119-127(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zgdjgcxb201305017 [4] 桂勇, 韩勤锴, 李峥, 等.风机行星齿轮系统齿轮裂纹故障诊断[J].振动、测试与诊断, 2016, 36(1):169-175 http://d.old.wanfangdata.com.cn/Periodical/zdcsyzd201601027Gui Y, Han Q K, Li Z, et al. The fault diagnosis of cracks in the planetary gear system of wind turbine[J]. Journal of Vibration, Measurement & Diagnosis, 2016, 36(1):169-175(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdcsyzd201601027 [5] 陈彬强, 张周锁, 訾艳阳, 等.机械故障诊断的衍生增强离散解析小波分析框架[J].机械工程学报, 2014, 50(17):77-86 http://d.old.wanfangdata.com.cn/Periodical/jxgcxb201417012Chen B Q, Zhang Z S, Zi Y Y, et al. Novel ensemble analytic discrete framelet expansion for machinery fault diagnosis[J]. Journal of Mechanical Engineering, 2014, 50(17):77-86(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jxgcxb201417012 [6] 于德介, 程军圣, 杨宇.Hilbert-Huang变换在齿轮故障诊断中的应用[J].机械工程学报, 2005, 41(6):102-107 doi: 10.3321/j.issn:0577-6686.2005.06.020Yu D J, Cheng J S, Yang Y. Application of Hilbert-Huang transform method to gear fault diagnosis[J]. Journal of Mechanical Engineering, 2005, 41(6):102-107(in Chinese) doi: 10.3321/j.issn:0577-6686.2005.06.020 [7] 李志农, 朱明, 褚福磊, 等.基于经验小波变换的机械故障诊断方法研究[J].仪器仪报, 2013, 35(11):2423-2432 http://d.old.wanfangdata.com.cn/Periodical/yqyb201411003Li Z N, Zhu M, Chu F L, et al. Mechanical fault diagnosis method based on empirical wavelet transform[J]. Chinese Journal of Scientific Instrument, 2014, 35(11):2423-2432(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/yqyb201411003 [8] Marron J S, Ramsay J O, Sangalli L M, et al. Functional data analysis of amplitude and phase variation[J]. Statistical Science, 2015, 30(4):468-484 doi: 10.1214/15-STS524 [9] 苏为华, 孙利荣, 崔峰.一种基于函数型数据的综合评价方法研究[J].统计研究, 2013, 30(2):88-94 doi: 10.3969/j.issn.1002-4565.2013.02.012Su W H, Sun L R, Cui F. One comprehensive evaluation method study based on functional data[J]. Statistical Research, 2013, 30(2):88-94(in Chinese) doi: 10.3969/j.issn.1002-4565.2013.02.012 [10] Hsing T, Eubank R. Theoretical foundations of functional data analysis, with an introduction to linear operators[M]. Chichester:John Wiley & Sons, 2015 [11] Petersen A, M?ller H G. Functional data analysis for density functions by transformation to a Hilbert space[J]. The Annals of Statistics, 2016, 44(1):183-218 doi: 10.1214/15-AOS1363 [12] Tarabelloni N, Ieva F. On data robustification in functional data analysis[C]//Affiliation: MOX-Modeling and Scientific Computing, Politecnico di Milano, 2016 [13] 冯志鹏, 赵镭镭, 褚福磊.行星齿轮箱齿轮局部故障振动频谱特征[J].中国电机工程学报, 2013, 33(5):119-127 http://d.old.wanfangdata.com.cn/Periodical/zgdjgcxb201305017Feng Z P, Zhao L L, Chu F L. Vibration spectral characteristics of localized gear fault of planetary gearboxes[J]. Proceedings of the CSEE, 2013, 33(5):119-127(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zgdjgcxb201305017 [14] Liu B, Xiao Y S, Cao L B, et al. SVDD-based outlier detection on uncertain data[J]. Knowledge and Information Systems, 2013, 34(3):597-618 doi: 10.1007/s10115-012-0484-y [15] Jiang Q C, Yan X F. Just-in-time reorganized PCA integrated with SVDD for chemical process monitoring[J]. AIChE Journal, 2014, 60(3):949-965 doi: 10.1002/aic.14335 [16] Li G N, Hu Y P, Chen H X, et al. An improved fault detection method for incipient centrifugal chiller faults using the PCA-R-SVDD algorithm[J]. Energy and Buildings, 2015, 116:104-113 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=203b25fa812cf3dcc0e7b511177fb096