Real-time Lifetime Prediction Method Based on Wavelet Support Vector Regression and Fuzzy c-means Clustering
-
摘要: 针对产品的性能退化轨迹呈现为非线性特性, 且个体的性能退化数据为小样本的情形, 为了充分利用同类产品的性能退化数据进行特定个体的实时寿命预测, 从研究退化轨迹相似性的角度出发, 提出一类基于小波支持向量回归机 (Wavelet support vector regression, WSVR)和模糊C均值(Fuzzy c-means, FCM)聚类的实时寿命预测方法. 该方法分为离线和实时两个阶段: 离线阶段先采用WSVR对同类产品的性能退化数据进行规范化处理, 接着对规范化数据进行FCM聚类, 然后,基于WSVR建立各聚类中心的退化轨迹模型;在实时阶段,针对特定个体的历史测量数据是否规范化,分别提出两种实时退 化轨迹建模和寿命预测方法——隶属度加权法和误差加权法. 最后, 通过两个实例分析验证了所提方法的有效性.Abstract: For the case where the products have nonlinear performance degradation paths and there is little performance degradation data for each individual, in order to take full advantage of performance degradation data of the same kind of products in individual real-time lifetime prediction, as viewed from the comparability of degradation paths, a class of real-time lifetime prediction methods are proposed, on the basis of wavelet support vector regression (WSVR) and fuzzy c-means (FCM) clustering . The methods consist of two stages, called offline and real-time stages. In the offline stage, WSVR is employed to normalize performance degradation data, and FCM is used to classify the normalized data, then WSVR is adopted again to build the cluster centers' degradation path models. In the real-time stage, depending on whether the specific individual's historical data is normalized or not, two real-time model-building methods, named degree-of-membership-based weighted method and error-based weighted method, are proposed. And then, by combining with the specific individual's real-time measurement, its degradation path model is renewed, and its lifetime is predicted. Two example analyses are executed based on fatigue crack growth data and the CG36A transistor degradation data, and the experimental results validate the validity of the two proposed methods.
-
[1] Hu Chang-Hua, Xu Hua-Long. Design and Analysis of Fault-Tolerant Control and Fault Diagnosis for Control System. Beijing: National Defense Industry Press, 2000(胡昌华, 许化龙. 控制系统故障诊断与容错控制的分析和设计. 北京: 国防工业出版社, 2000)[2] Si X S, Wang W, Hu C H, Zhou D H. Remaining useful life estimation —— a review on the statistical data driven approaches. European Journal of Operational Research, 2011, 213(1): 1-14[3] Gebraeel N Z, Lawley M A, Liu R, Parmeshwaran V. Residual life predictions from vibration-based degradation signals: a neural network approach. IEEE Transactions on Industrial Electronics, 2004, 51(3): 694-700[4] Lu J C, Meeker W Q. Using degradation measures to estimate a time-to-failure distribution. Technometrics, 1993, 35(2): 161-174[5] Chan V. Degradation-Based Reliability in Outdoor Environments [Ph.D. dissertation], Iowa State University, USA, 2001[6] Lim K E, Baek J S, Lee E Y. A random shock model for a continuously deteriorating system. Journal of Quality in Maintenance Engineering, 2005, 11(3): 206-215[7] Ramakrishnan A, Pecht M. A life consumption monitoring methodology for electronic systems. IEEE Transactions on Components and Packaging Technologies, 2003, 26(3): 625-634[8] Son K I, Soma M. Dynamic life-estimation of CMOS ICs in real operating environment: precise electrical method and MLE. IEEE Transactions on Reliability, 1997, 46(1): 31-37[9] Lv Ke-Hong, Qiu Jing, Liu Guan-Jun. Research on life prognosis method for electronics based on dynamic damage and optimization AR model. Acta Armamentarii, 2009, 30(1): 91-95(吕克洪, 邱静, 刘冠军. 基于动态损伤及优化AR模型的电子器件寿命预测方法研究. 兵工学报, 2009, 30(1): 91-95)[10] Wu S J, Tsai T R. Estimation of time-to-failure distribution derived from a degradation model using fuzzy clustering. Quality and Reliability Engineering International, 2000, 16(4): 261-267[11] Chinnam R B. On-line reliability estimation for individual components using statistical degradation signal models. Quality and Reliability Engineering International, 2002, 18(1): 53-73[12] Gebraeel N Z, Lawley M A, Li R, Ryan J K. Residual-life distributions from component degradation signals: a Bayesian approach. IIE Transactions, 2005, 37(6): 543-557[13] Gebraeel N Z. Sensory-updated residual life distributions for components with exponential degradation patterns. IEEE Transactions on Automation Science and Engineering, 2006, 3(4): 382-393[14] Gebraeel N Z, Elwany A, Pan J. Residual life predictions in the absence of prior degradation knowledge. IEEE Transactions on Reliability, 2009, 58(1): 106-117[15] Chen Z H, Zheng S R. Lifetime distribution based degradation analysis. IEEE Transactions on Reliability, 2005, 54(1): 3-10[16] Deng Ai-Min, Chen Xun, Zhang Chun-Hua, Wang Ya-Shun. Reliability assessment based on performance degradation data. Journal of Astronautics, 2006, 27(3): 546-552(邓爱民, 陈循, 张春华, 汪亚顺. 基于性能退化数据的可靠性评估. 宇航学报, 2006, 27(3): 546-552)[17] Zhuang Dong-Chen. Degradation Failure Model and Statistical Analysis [Ph.D. dissertation], East China Normal University, China, 1994(庄东辰. 退化失效模型及其统计分析 [博士学位论文], 华东师范大学, 中国, 1994)[18] Xu Z G, Zhou D H. A degradation measurements based real-time reliability prediction method. In: Proceedings of the 6th IFAC Symposium on Fault Detection, Supervision and Safety of Technical Processes. Beijing, China: IFAC, 2006. 950-955[19] Lu S, Lu H, Kolarik W J. Multivariate performance reliability prediction in real-time. Reliability Engineering and System Safety, 2001, 72(1): 39-45[20] You Qi, Zhao Yu, Ma Xiao-Bing. Performance reliability assessment for products based on time series analysis. Journal of Beijing University of Aeronautics and Astronautics, 2009, 35(5): 644-648(尤琦, 赵宇, 马小兵. 产品性能可靠性评估的时序分析方法. 北京航空航天大学学报, 2009, 35(5): 644-648)[21] Xu Z G, Ji Y D, Zhou D H. Real-time reliability prediction for a dynamic system based on the hidden degradation process identification. IEEE Transactions on Reliability, 2008, 57(2): 230-242[22] Huang R Q, Xi L F, Li X L, Liu C R, Qiu H, Lee J. Residual life predictions for ball bearings based on self-organizing map and back propagation neural network methods. Mechanical Systems and Signal Processing, 2007, 21(1): 193-207[23] Gebraeel N Z, Lawley M A. A neural network degradation model for computing and updating residual life distributions. IEEE Transactions on Automation Science and Engineering, 2008, 5(1): 154-163[24] Hu Chang-Hua, Hu Jin-Tao, Zhang Wei, Ping Zhen-Hai. Reliability assessment of performance degradation using support vector machines. Systems Engineering and Electronics, 2009, 31(5): 1246-1249(胡昌华, 胡锦涛, 张伟, 平振海. 支持向量机用于性能退化的可靠性评估. 系统工程与电子技术, 2009, 31(5): 1246-1249)[25] Wu J, Deng C, Shao X Y, Xie S Q. A reliability assessment method based on support vector machines for CNC equipment. Science in China Series E: Technological Sciences, 2009, 52(7): 1849-1857[26] Li Chang-You, Xu Min-Qiang, Guo Song, Wang Ri-Xin, Gao Jing-Bo. Real-time reliability assessment based on gamma process and Bayesian estimation. Journal of Astronautics, 2009, 30(4): 1715-1719(李常有, 徐敏强, 郭耸, 王日新, 高晶波. 基于Gamma过程及贝叶斯估计的实时可靠性评估. 宇航学报, 2009, 30(4): 1715-1719)[27] Zhao Zhao, Dong Dou-Dou, Zhou Jing-Lun, Feng Jing. Real-time reliability evaluation for degradation failure product based on Bayes method. Journal of National University of Defense Technology, 2007, 29(6): 116-120(赵炤, 董豆豆, 周经伦, 冯静. 基于贝叶斯方法的退化失效型产品实时可靠性评估. 国防科技大学学报, 2007, 29(6): 116-120)[28] Deng Nai-Yang, Tian Ying-Jie. A New Method of Data Mining: Support Vector Machine. Beijing: Science Press, 2004(邓乃杨, 田英杰. 数据挖掘中的新方法: 支持向量机. 北京: 科学出版社, 2004)[29] Zhang Li, Zhou Wei-Da, Jiao Li-Cheng. Wavelet kernel function network. Journal of Infrared Millimeter Waves, 2001, 20(3): 223-227(张莉, 周伟达, 焦李成. 子波核函数网络. 红外与毫米波学报, 2001, 20(3): 223-227)[30] Gao Xin-Bo. Fuzzy Cluster Analysis and Its Applications. Xi'an: Xi'an University of Electronic Science and Technology Press, 2004(高新波. 模糊聚类分析及其应用. 西安: 西安电子科技大学出版社, 2004)
点击查看大图
计量
- 文章访问数: 2087
- HTML全文浏览量: 48
- PDF下载量: 1115
- 被引次数: 0