首页|基于指数扩散过程的退化模型误指定分析

基于指数扩散过程的退化模型误指定分析

扫码查看
退化模型是对性能退化产品进行可靠性分析的基础,鉴于误用模型直接影响可靠性评估的精度,研究将适用范围广泛的指数扩散过程模型误指定为3种常见的维纳过程模型、伽马过程模型和逆高斯过程模型,通过引入相对偏差和相对变异率两个指标,以量化模型误指定对产品平均寿命的影响程度,并采用GaAs激光器数据和蒙特卡罗方法进行实例验证.结果表明,模型误指定将对产品平均寿命估计带来较大影响,在性能退化产品的可靠性研究中应尽量选取适用范围更广的指数扩散过程模型,以提高可靠性评估的准确性.
Mis-specification analysis of degradation model based on exponential dispersion process
Degradation model is the basis for reliability analysis of degraded products.Owing to the problem that the mis-specification of degradation model directly affects the accuracy of reliability assessment,the issue of mis-specif-ying the widely applicable exponential dispersion process model as three commonly used Wiener process model,gamma process model and inverse Gaussian process model was studied.By calculating the relative bias and relative variability,the influence of the model mis-specification was analyzed,and the GaAs laser data was select for exam-ple verification.The results showed that the model mis-specification would have a greater influence on the estimation of the products'average life.Therefore,in the reliability consideration of degraded products,exponential dispersion process model with wider application scope should be selected for modeling analysis.

model mis-specificationexponential dispersion processmean-time-to-failurequasi maximum likelihood estimation

鄢伟安、李欣忆、张士杰、刘卫东

展开 >

南昌大学先进制造学院,江西 南昌 330031

中国科学院大学经济与管理学院,北京 100190

华东交通大学交通运输工程学院,江西 南昌 330013

模型误指定 指数扩散过程 平均失效时间 伪极大似然估计

国家自然科学基金资助项目国家自然科学基金资助项目江西省研究生创新专项资助项目

7186101172071099YC2020-S332

2024

计算机集成制造系统
中国兵器工业集团第210研究所

计算机集成制造系统

CSTPCD北大核心
影响因子:1.092
ISSN:1006-5911
年,卷(期):2024.30(10)