基于非线性漂移Wiener过程的产品实时可靠性评估  

Real-time reliability evaluation for product with nonlinear drift-based Wiener process

作　　者：王小林[1] 郭波[1] 程志君[1] 

机构地区：[1]国防科技大学信息系统与管理学院,湖南长沙410073

出　　处：《中南大学学报（自然科学版）》2013年第8期3203-3209,共7页Journal of Central South University:Science and Technology

基　　金：国家自然科学基金资助项目(60904002);国防科技大学预研基金资助项目(JC12 02 05)

摘　　要：提出基于非线性维纳退化过程评估单个产品实时可靠性的方法。该方法能够有效融合先验信息和产品自身的退化信息。首先,采用非线性维纳过程描述产品性能退化的非线性特征,产品之间的差异通过随机化参数表示。然后,利用两阶段方法估计模型中的未知参数。当获得新的退化数据信息时,根据贝叶斯方法对参数进行更新,进一步实现对产品实时可靠性的评估。最后,通过一个仿真示例以及电容器实例验证文中方法的有效性。研究结果表明:与线性维纳过程相比,非线性维纳过程能够较好地描述产品性能退化的非线性,并能够给出可信的评估结果。With respect to the Wiener degradation process with a nonlinear drift, a real-time reliability evaluation method for a single product was proposed. This method could effectively fuse the prior information and the product＇s own degradation information. Firstly, the nonlinearity of the performance degradation was characterized by a nonlinear drift-based Wiener process and the product-to-product variability was described by random parameters. Then, unknown parameters were estimated by using the two-stage method. When new degradation information was available, parameters were updated by using the Bayesian method. After that, the real-time reliability was further evaluated. Finally, the validity of the proposed method is demonstrated by a simulation case and a capacitor case. Results show that the nonlinearity of the performance degradation can be effectively captured by the nonlinear drift-based Wiener process and reliable evaluation results can be obtained, compared with the linear drift-based Wiener process.

关 键 词：非线性漂移维纳过程 实时可靠性评估 贝叶斯方法 似然比检验 

分 类 号：TB114.3[理学—概率论与数理统计]