Please wait a minute...
工程设计学报  2024, Vol. 31 Issue (1): 59-66    DOI: 10.3785/j.issn.1006-754X.2024.03.304
可靠性与保质设计     
无失效威布尔情形下基于双修正多层Bayes的可靠性评估
龙足腾1(),郑波1,2(),甯洋1,罗金超1
1.中国民用航空飞行学院 航空电子电气学院,四川 广汉 618307
2.核工业西南物理研究院,四川 成都 610225
Reliability estimation based on double-modified hierarchical Bayes in the zero-failure Weibull case
Zuteng LONG1(),Bo ZHENG1,2(),Yang NING1,Jinchao LUO1
1.Institute of Electronic and Electrical Engineering, Civil Aviation Flight University of China, Guanghan 618307, China
2.Southwestern Institute of Physics, Chengdu 610225, China
 全文: PDF(2368 KB)   HTML
摘要:

利用配分布曲线法构建无失效数据的可靠性评估模型时,往往采用E-Bayes或多层Bayes来估计失效概率,但因其估计能力有限,整体可靠度点估计精度不高。针对威布尔分布数据提出了一种新的双修正多层Bayes方法,用于改善失效概率估计,并完成可靠度点估计。在多层Bayes的基础上,通过修正失效概率上下限值来减小失效概率估计误差,然后结合加权最小二乘法和威布尔分布可靠度函数确定参数估计值和可靠度曲线,从而得到可靠度点估计。利用Monte-Carlo仿真试验,发现新方法能将参数估计相对误差控制在10%以下,并且得到的可靠度值更加趋近真值。通过实例分析发现,在超参数c取不同值的情况下,采用新方法得到的可靠度都更加接近工程实际。新方法适用性分析表明,形状参数β是影响估计精度的关键因素,且在β>2.2时新方法具有明显的优势及良好的稳健性。研究结果可为其他寿命分布无失效数据的可靠性评估提供参考。

关键词: 无失效威布尔失效概率估计双修正多层Bayes可靠度估计适用性分析    
Abstract:

Hierarchical Bayes or E-Bayes is frequently used to estimate the failure probability when building a zero-failure reliability estimation model utilizing the concept of a distribution curve. The overall accuracy of reliability point estimation is not high bacause of the limited estimation ability. A new double-modified hierarchical Bayes method was proposed for the Weibull distribution data to improve the failure probability estimation and accomplish reliability point estimation. On the basis of hierarchical Bayes, the failure probability estimation error was reduced by correcting the upper and lower limits of the failure probability. Combining weighted least squares method and Weibull distribution reliability function determined parameter estimates and reliability curves, thereby obtaining reliability point estimates. Using Monte-Carlo simulation test, the new method could control the relative error of parameter estimation below 10%, and the obtained reliability was closer to the true value. Through the example analysis, the reliability obtained by the new method was closer to the engineering reality when the hyperparameterctook different values. The applicability analysis of the new method showed that the shape parameter β was the key factor affecting the estimation accuracy. The new method had obvious advantages and good robustness when β>2.2. The results of the study can provide a reference for the reliability assessment of other life distribution with zero-failure data.

Key words: zero-failure and Weibull    failure probability evaluation    double-modified hierarchical Bayes    reliability evaluation    applicability analysis
收稿日期: 2023-10-28 出版日期: 2024-03-04
CLC:  TB 114.3  
基金资助: 中国民用航空局民航安全能力建设资金资助项目(MHAQ2022004)
通讯作者: 郑波     E-mail: ttengzulong@163.com;bzheng@cafuc.edu.cn
作者简介: 龙足腾(2000—),男,四川资阳人,硕士生,从事无失效情形的可靠性评估方法研究,E-mail: ttengzulong@163.com, https://orcid.org/0009-0009-3690-7345
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
龙足腾
郑波
甯洋
罗金超

引用本文:

龙足腾,郑波,甯洋,罗金超. 无失效威布尔情形下基于双修正多层Bayes的可靠性评估[J]. 工程设计学报, 2024, 31(1): 59-66.

Zuteng LONG,Bo ZHENG,Yang NING,Jinchao LUO. Reliability estimation based on double-modified hierarchical Bayes in the zero-failure Weibull case[J]. Chinese Journal of Engineering Design, 2024, 31(1): 59-66.

链接本文:

https://www.zjujournals.com/gcsjxb/CN/10.3785/j.issn.1006-754X.2024.03.304        https://www.zjujournals.com/gcsjxb/CN/Y2024/V31/I1/59

组号i截尾时间ti样本数ni未失效数si
1t1n1s1
2t2n2s2
????
ktknksk
表1  无失效数据结构
图1  形状参数β对威布尔函数的影响
组号i截尾时间ti/h样本数ni未失效数si
1273728
2558621
3709515
4851410
593636
61 04923
71 14111
表2  仿真的无失效数据
截尾时间ti/h真值传统BayesE-Bayes多层Bayes

本文

方法

2730.000 60.033 30.031 30.030 90.007 3
5580.011 60.043 50.040 10.039 40.014 8
7090.023 20.058 80.052 80.051 80.024 2
8510.034 00.083 30.071 90.070 10.038 3
9360.048 90.125 00.101 40.098 10.063 8
1 0490.057 20.200 00.146 90.141 50.120 7
1 1410.070 60.333 30.211 80.120 70.279 1
表3  不同方法下的失效概率估计值
方法βηEmr-βEmr-ηc
传统Bayes2.082 336.60.307 00.065 4-
E-Bayes1.683 434.90.439 10.373 95
多层Bayes1.663 568.70.447 10.427 55
本文方法2.822 333.90.061 20.066 55
表4  不同方法下的参数估计及其相对误差
图2  不同方法下的可靠度
组数i截尾时间ti/h样本数ni未失效数si
1135621
2280515
3370410
466536
51 15023
61 30011
表5  某轴承的无失效数据
图3  在c=4~7时轴承的可靠度
图4  适用性分析的试验流程
图5  各方法估计β的相对误差
图6  各方法估计η的相对误差
1 MARTZ JR H F, WALLER R A. A Bayesian zero-failure (BAZE) reliability demonstration testing procedure[J]. Journal Of Quality Technology, 1979, 11(3): 128-138.
2 茆诗松, 罗朝斌. 无失效数据的可靠性分析[J]. 数理统计与应用概率, 1989, 4(4): 489-506.
MAO S S, LUO C B. Reliability analysis of zero-failure data[J]. Mathematical Theory and Application Probability, 1989, 4(4): 489-506.
3 茆诗松, 夏剑峰. 无失效数据的多层贝叶斯分析[J]. 高校应用数学学报: A辑, 1992, 7(3): 11.
MAO S S, XIA J F. Hierarchical bayesian analysis of zero-failure data[J]. Applied Mathematics A Journal of Chinese Universities: Series A, 1992, 7(3): 11.
4 韩明. 多层先验分布的构造及其应用[J]. 运筹与管理, 1997, 6(3): 10.
HAN M. The structure of hierarchical prior distribution and it’s applications[J]. Operations Research and Management Science, 1997, 6(3): 10.
5 韩明. 产品无失效数据的综合处理[J]. 机械工程学报, 2003, 39(2): 4. doi:10.3901/jme.2003.02.129
HAN M. Synthesized process for zero-failure data of the products[J]. Journal of Mechanical Engineering, 2003, 39(2): 4.
doi: 10.3901/jme.2003.02.129
6 蔡国梁, 吴来林, 唐晓芬. 双超参数无失效数据的E-Bayes可靠性分析[J]. 江苏大学学报(自然科学版), 2010, 31(6): 4.
CAI G L, WU L L, TANG X F. E-bayes reliability analysis of double hyperparameters zero failure data[J]. Journal of Jiangsu University(Natural Science Edition), 2010, 31(6): 4.
7 王璨, 董玉革, 黄聿锋, 等. 基于无失效数据失效概率多层Bayes估计的单失效数据可靠性分析[J]. 机械工程师, 2019(9): 4.
WANG C, DONG Y G, HUANG Y F, et al. Reliability analysis of single failure datum based on muti-layer bayes estimation[J]. Mechanical Engineer, 2019(9): 4.
8 YIN Y C, HUANG H Z, PENG W, et al. An E-bayesian method for reliability analysis of exponentially distributed products with zero-failure data[J]. Eksploatacja i Niezawodnosc - Maintenance and Reliability, 2016, 18(3): 445-449.
9 宁江凡, 鄢小清, 张士峰. 液体火箭发动机无失效条件下的可靠性分析方法[J].国防科技大学学报, 2006, 28(5): 4. doi:10.3969/j.issn.1001-2486.2006.05.005
NING J F, YAN X Q, ZHANG S F. Study on reliability analysis method of liquid rocket engine in the case of zero-failure[J]. Journal of National University of Defense Technology, 2006, 28(5): 4.
doi: 10.3969/j.issn.1001-2486.2006.05.005
10 ERTO P, GIORGIO M. Modified ‘practical bayes-estimators’ reliability theory[J]. IEEE Transactions on Reliability, 1996, 45(1): 132-137.
11 KWON Y I. Design of bayesian zero-failure reliability demonstration test and its application[J]. Journal of Applied Reliability, 2013, 13(1): 1-10.
12 刘海涛, 张志华. 威布尔分布无失效数据的Bayes可靠性分析[J]. 系统工程理论与实践, 2008, 28(11): 6. doi:10.3321/j.issn:1000-6788.2008.11.016
LIU H T, ZHANG Z H. Bayesian reliability analysis of weibull zero failure data[J]. Systems Engineering-Theory & Practice, 2008, 28(11): 6.
doi: 10.3321/j.issn:1000-6788.2008.11.016
13 JIANG P, LIM J H, ZUO M J, et al. Reliability estimation in a Weibull lifetime distribution with zero‐failure field data[J]. Quality and Reliability Engineering International, 2010, 26(7): 691-701.
14 ZHANG C W. Weibull parameter estimation and reliability analysis with zero-failure data from high-quality products[J]. Reliability Engineering & System Safety, 2021, 207(3): 107321.
15 徐天群, 陈跃鹏, 徐天河, 等. 威布尔分布场合下无失效数据的可靠性参数估计[J]. 统计与决策, 2012(13): 4.
XU T Q, CHEN Y P, XU T H, et al. Estimation of reliability parameters for zero-failure data under Weibull distribution[J]. Statistics & Decision, 2012(13): 4.
16 高攀东, 沈雪瑾, 陈晓阳, 等. 无失效数据下航空轴承的可靠性分析[J]. 航空动力学报, 2015, 30(8): 8.
GAO P D, SHEN X J, CHEN X Y, et al. Reliability analysis for aircraft bearing with zero-failure data[J]. Journal of Aerospace Power, 2015, 30(8): 8.
17 李海洋, 谢里阳, 刘杰, 等. 无失效数据场合智能换刀机器人中轴承的可靠性评估[J]. 机械工程学报, 2019, 55(2): 186-194. doi:10.3901/jme.2019.02.186
LI H Y, XIE L Y, LIU J, et al. Reliability evaluation of bearing in the intelligent robot for changing the hob without failure data[J]. Journal of Mechanical Engineering, 2019, 55(2): 186-194.
doi: 10.3901/jme.2019.02.186
18 LI H Y, XIE L Y, LI M, et al. Reliability assessment of high-quality and long-life products based on zero-failure data[J]. Quality and Reliability Engineering, 2019, 35(1): 470-482.
19 王瑞祥, 许凌天, 陈晓阳, 等. 小样本无失效寿命试验数据的轴承可靠性评估[J].航空动力学报, 2021, 36(11): 2400-2409.
WANG R X, XU L T, CHEN X Y, et al. Reliability evaluation of bearing based on small samples and zero-failure life test data[J]. Journal of Aerospace Power, 2021, 36(11): 2400-2409.
20 CHEN R, CHEN G, LIU X, et al. Reliability prediction method for low-cycle fatigue life of compressor disks based on the fusion of simulation and zero-failure data[J]. Applied Sciences, 2022, 12(9): 4318.
21 李爽, 李云飞, 高海军. 威布尔分布无失效数据场合失效概率估计的改进方法[J]. 统计与决策, 2019(24): 4.
LI S, LI Y F, GAO H J. An improved method for estimation of failure probability in weibull distribution without failure date[J]. Statistics & Decision, 2019(24): 4.
[1] 马天政, 秦代成, 张义民. 基于遗传算法的谐波齿轮可靠性优化设计[J]. 工程设计学报, 2011, 18(4): 246-250.
[2] 林瑞进, 陈文华, 刘娟, 高亮, 潘骏. 航天电连接器加速性能退化试验可行性研究[J]. 工程设计学报, 2010, 17(4): 317-320.
[3] 高明君, 郝建平, 刘继民, 曲长征. 基于系统工程过程的维修性设计要求生成过程研究[J]. 工程设计学报, 2010, 17(2): 81-85.