基于HMM和优化的PF的数控转台精度衰退模型

摘要
图/表
参考文献(18)
相关文章 (3)

全文: PDF (2495 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要

针对数控转台精度衰退状态缺乏有效的评估方法的问题，提出一种数控转台重复定位精度衰退趋势预测模型，该模型结合了隐马尔科夫（Hidden Markov Model, HMM）算法和粒子滤波（Particle Filtering, PF）算法，其中粒子滤波算法使用粒子群算法（Particle Swarm Optimization, PSO）优化了初始参数。选择了从数控转台精度衰退加速寿命试验中获得的振动信号作为研究数据。首先，通过聚合经验模态与主成分分析（EEMD-PCA）算法对原始信号降噪,并提取含有故障特征的信号进行信号重构；然后，使用统计特征量作为观察值训练获得HMM模型，对数控转台精度衰减做出早期诊断，并由此获得数控转台精度健康状态指标；最后，使用粒子滤波算法建立数控转台精度衰退预测模型，并预测精度的剩余寿命。在以第50组数据为预测起始点时，预测的剩余寿命为21，实际测量的结果为17，相差4，比较接近。综合分析模型计算与试验测量的结果表明，该模型可有效地预测数控转台精度的变化趋势和剩余寿命。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	王刚1
	陈捷1
	2
	洪荣晶1
	2
	王华1
	2

关键词 ：数控转台')" href="#">

数控转台, 隐马尔科夫模型, 粒子滤波算法, 定位精度, 剩余寿命

Abstract：A novel prediction approach for NC rotary tables was proposed based on the hidden Markov model(HMM) and the particle filtering(PF) to estimate the degradation trend of the replicated positional accuracy. The initial parameter of particle filtering was optimized by the particle swarm optimization(PSO). The vibration signal was selected as the data for research, which was obtained from an accelerated accuracy degradation test of a NC rotary table. The original signal was denoised and reconstructed by an ensemble empirical mode decomposition and principal component analysis. Then, a HMM was trained by an observation matrix which was composed of the statistical characteristic values, and the diagnosis of early positional accuracy degradation and the health status indicators of the accuracy were obtained. Finally, the degradation trend model of the positional accuracy was established by the particle filtering, and the residual accuracy life was calculated.When fiftieth sets of data were used as the starting point of prediction, the predicted residual life is 21, and the actual measurement result is 17, which are close to each other. Comparing the results of model calculations and experimental measurements, it is shown that the approach is efficient to estimate the degradation trend of the positional accuracy and the residual accuracy life.

Key words： NC rotary table Hidden Markov Model Particle Filtering positional accuracy residual accuracy life

收稿日期: 2017-08-21 出版日期: 2018-03-15

引用本文:

王刚1，陈捷1,2，洪荣晶1,2，王华1,2. 基于HMM和优化的PF的数控转台精度衰退模型[J]. 振动与冲击, 2018, 37(6): 7-13.
WANG Gang1, CHEN Jie1，2, HONG Rong-jing1，2, WANG Hua1，2. Model for the positional accuracy degradation of NC rotary tables based on the hidden Markov model and optimized particle filtering. JOURNAL OF VIBRATION AND SHOCK, 2018, 37(6): 7-13.

链接本文:

http://jvs.sjtu.edu.cn/CN/ 或 http://jvs.sjtu.edu.cn/CN/Y2018/V37/I6/7

[1] YU C J, HUANG X D, FANG C G, et al. Research on Damping and Vibration Characteristic of the Large and Precision NC Rotary Table[J]. Advanced Materials Research, 2010, 97-101:1216-1222.
[2] ŽVOKELJ M, ZUPAN S, PREBIL I. Non-linear multivariate and multiscale monitoring and signal denoising strategy using Kernel Principal Component Analysis combined with Ensemble Empirical Mode Decomposition method [J]. Mechanical Systems & Signal Processing, 2011, 25(7):2631-2653.
[3] YU CJ, HUANG X D, FANG C G. Research on dynamic characteristics of NC rotary table considering leakage factors of its hydrostatic guideway[J]. ARCHIVE Proceedings of the Institution of Mechanical Engineers Part C Journal of Mechanical Engineering Science 1989-1996 (vols 203-210), 2012, 226(11):2674-2685.
[4] HENG A, ZHANG S, TAN A C C, et al. Rotating machinery prognostics: State of the art, challenges and opportunities [J]. Mechanical Systems & Signal Processing, 2009, 23(3):724-739.
[5] T. BENKEDJOUH, K. MEDJAHER, N. ZERHOUNI. Remaining useful life estimation based on nonlinear feature reduction and support vector regression [J].Engineering Applications of Artificial Intelligence, 2013, 26:1751-1760.
[6] LU C, Chen J, Hong R, et al. Degradation trend estimation of slewing bearing based on LSSVM model [J]. Mechanical Systems & Signal Processing, 2016, 76–77:353-366.
[7] MATEJ Z., SAMO Z., IVAN P. Non-linear multivariate and multiscale monitoring and signal denoising strategy using kernel principal component analysis combined with ensemble empirical mode decomposition method. Mechanical Systems and Signal Processing, Vol. 25, 2011, p. 2631-2653.
[8] FENG Y, HUANG X, HONG R, et al. Residual useful life prediction of large-size low-speed slewing bearings - a data driven method[J]. Journal of Vibroengineering, 2015, 17(8):4164-4179.
[9] 于春健.多物理场耦合的大重型数控回转工作台动静态精度研究[D].南京：南京工业大学，2012.
YU C J. Research on Dynamic and Static Indexing Accuracy of the Large and Heavy NC Rotary Table with Multiphysics Coupling [D].Nan Jing: Nanjing Tech University, 2012(in Chinese).
[10] WANG D, MIAO Q, ZHOU Q, et al. An Intelligent Prognostic System for Gear Performance Degradation Assessment and Remaining Useful Life Estimation [J]. Journal of Vibration & Acoustics, 2015, 137(2):12.
[11] 张西宁,雷威,李兵. 主分量分析和隐马尔科夫模型结合的轴承监测诊断方法[J]. 西安交通大学学报,2017,(06).
ZHANG Xining, LEI Wei, LI Bing. Bearing Fault Detection and Diagnosis Method Based on Principal Component Analysis and Hidden Markov Model [J].Journal of Xi'an Jiaotong University,2017,(06)(in Chinese).
[12] 余龙华,王宏,钟洪声. 基于隐马尔科夫模型的人脸识别[J]. 计算机技术与发展,2012,(02):25-28.
YU Long-hua，WANG Hong，ZHONG Hong-sheng. Face Recognition Based on Hidden Markov Model [J].Computer Technology and Development, 2012, (02):25-28(in Chinese).
[13] RABINER L R. A tutorial on hidden Markov models and selected applications in speech recognition [J]. Readings in Speech Recognition, 1990, 77(2):267-296.
[14] ARULAMPALAM, M.S., MASKELL, S., Gordon, N. & Clapp, T. (2002). A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking. IEEE Transactions on Signal Processing. 50(2):174-188.
[15] 王法胜,鲁明羽,赵清杰,袁泽剑. 粒子滤波算法[J]. 计算机学报,2014,(08):1679-1694.
WANG Fa-Sheng, LU Ming-Yu, ZHAO Qing-Jie, et al. Particle Filtering Algorithm [J]. Chinese Journal of computer, 2014, (08):1679-1694(in Chinese).
[16] 方正,佟国峰,徐心和. 粒子群优化粒子滤波方法[J]. 控制与决策,2007,(03):273-277.
FANG Zheng, TONG Guo-feng, XU Xin-he. Particle swarm optimized particle filter [J].Control and Decision, 2007, (03):273-277(in Chinese).
[17] 王维博,林川,郑永康. 粒子群算法中参数的实验与分析[J]. 西华大学学报(自然科学版),2008,(01):76-80+105-106.
WANG Wei-bo, LIN Chuan, et al.Experiment and Analysis of Parameters in Particle Swarm Optimization [J]. Journal of Xi Hua University,2008,(01):76-80+105-106(in Chinese).
[18] 封杨,黄筱调,陈捷,王华,洪荣晶. 大型回转支承非平稳振动信号的EEMD-PCA降噪方法[J]. 南京工业大学学报(自然科学版),2015,(03):61-66.
FENG Yang,HUANG Xiaodiao,CHEN Jie, et al. EEMD-PCA based denosing method for non-stationary vibration signals of large-size slewing bearings [J].Journal of Nanjing Tech University(Natural Science Edition) ,2015,(03):61-66(in Chinese).