用增长模型跟踪航天在研产品的可靠性_英文_

用增长模型跟踪航天在研产品的可靠性_英文_ ?ORIGINAL R ES EARCH?Tracking Rel ia bil ity f or Space Ca bin2borne Equipment in De2 velopment by Cro w Model CHEN J in2dun , J IAO Shu2jin , SUN Hai2long ( )Institute of Space Medico2Engineering , Beijing 100094 , China Abstract :Objective To study and track the reliability growth of manned spaceflight cabin2borne equipment in the course of its development . Method A new technique of reliability growth estimation and prediction , which is com2 ( ) posed of the Crow model and test data conversion TDCmethod was used. Result The estimation and prediction value of the reliability growth conformed to its expectations. Conclusion The method could dynamically estimate and predict the reliability of the equipment by making full use of various test information in the course of its develop2 ment . It offered not only a possibility of tracking the equipment reliability growth , but also the reference for quality control in manned spaceflight cabin2borne equipment design and development process. Key words :space cabin2borne equipment ;reliability growth ; Crow2model ; reliability estimation ; reliability tests ( ) 用增长模型跟踪航天在研产品的可靠性 . , 焦蜀晋 ,孙海龙. 航天医学与医学 ,2001 ,14 6: 391,394 ( ) 摘要 :目的 探讨一种跟踪航天在研产品可靠性增长变化情况的技术 。方法 应用克劳 Crow增长模型和 试验数据折算技术相结合的方法 。结果 该方法能够利用产品研制过程中的各种试验信息来动态地评 估和预测航天研制中产品的可靠性 ,跟踪结果与预期相符 。结论 该方法简单有效 ,可作为航天研制中 产品可靠性跟踪管理的工具 ,有推广价值 。 关键词 :航天产品 ;可靠性增长 ;Crow 模型 ;可靠性评估 ;可靠性试验 () 文章编号 :100220837 20010620391204 中图分类号 :V57文献标识码 :A 作者地址 :陈金盾. 航天医学工程研究所 ,北京 100094 For a manned spaceflight cabin2borne e2 MTBF or tion approach can dynamically estimate quipment the consequences of failure are so se2 other reliability parameters for the final system vere that much higher reliability is required. A design by fully using the prototype data from newly designed equipment should be developed different environmental condi2 various test in 6 from the prototype to the final production items , the Crow model and the TDC tions. First , to ensure that the requirements are met . There is technique were summarized in the paper . Sec2 usually a test2fix2test2fix procedure in which ond , the application of the combined approach prototypes may be built , tested and analyzed for in a manned spaceflight cabin2borne electronic failure modes to improve reliability through re2 equipment during the course of its design and design. It is important to use the prototype test development were described. data to track the reliability growth of the equip2 ment in the detailed design and development Method 1 ,2 phase. The Crow model can be used , it is a 3 ,5 Tracking the reliability of an equipment world wide used reliability growth model . means reliability growth analysis ,estimation and The test data may include performance test , prediction using test data during the course of its bench test , reliability test , accelerated life test , detailed design and development . It involves re2 or field test results. Each test has different en2 1 ,2 liability growth models and the test data conver2 vironment conditions. For this reason , the sion technique . The Crow model is probably the ( ) TDCtechnique was com2test data conversion most widely used model of the growth models on bined into the Crow model . The combina2 which the approach of monitoring reliability growth in the paper is based. Received date :2000211227 [ 3] The Cro w modelThe Crow model is the ratio between the time of B EC and the time intended for an item developed in a succession of models , which has a number of components of test environmental conditions is referred to as ( ) redesigned on the basis of failures of items of the environmental conversion factor ECF of the 2 K represent the test conditions. We usually let previous design. The unknown Poisson failure ECF. In practice , the value of K can be deter2 rate for a new model is assumed to remain con2 mined by knowledgeable and experienced ana2 stant throughout the existence of the model , but 3 ,4 lysts. For example , if the effectiveness of a each model has its own rate. A Weibull in2 stress test 1 h is equal to the B EC test 2 h , then tensity function is used : b - 1K = 2. Therefore , test data could be converted ()λ( ) 1 t= abt according to K for reliability estimation.λ ( ) where : t = failure rate after t operating Suppose that the reliability growth test pro2 hours , a > 0 , scale parameter , b > 0 , shape pa2 cedure of an equipment includes d stages of rameter . various stress tests as shown in Fig. 1. The first When b < 1 , reliability is increasing. stage starts at Tand stop s at T, and ECF is 0 1 When b > 1 , reliability is decreasing. If b = 1 , K. The beginning and the end of j stage time 1 the Weibull function is the exponential function is Tand T, and ECF is also K, where j = and the failure rate has no change . j - 1 j j In practice ,the test results can be used to 1 ,2 , , d . Then time of the beginning and the estimate the parameters of the Crow model with end of test stages is converted as4 ,7 statistical methods. T′= T 0 0 The test data conversion technique ( ()) T′= T′+ KT- j = 1 ,2 , ,d 2T j j - 1 j j j - 1Reliability is defined as the measurable capabil2 ity of an object to perform its intended function in the required time under specified conditions. Fig. 1 Relia bility gro wth test composed of various tests In practice , the reliability is estimated under the specified environmental conditions. Here , Failure times tin the whole growth test i the specified environmental conditions are re2 can be calculated as ferred to as basic environmental conditions ( )Tt′= T′+ Kt-, T?t?T, j - 1 j - 1 i j i j - 1 j i () B EC. However , there is a great difference on ()3 environmental conditions among performance j = 1 , , d i = 1 , , n test , environmental test , reliability test and Correspondingly, conversed time of the growth test end is T′. d matching test in the equipment development . Test data from them must be converted to the For example , a prototype of electronic e2 B EC before the reliability is estimated with Crow quipment for manned space flight underwent model by using these data . performance test , integrated test , environmental To achieve equal effectiveness of test , stress test , accelerated life test and matching Ta ble 1 Test data of the space ca bin2borne electronic equipment in prototype development ( )( )No . test item environmental factor test time h number of failures test time to failure h 1 performance test 0 . 5 360 6 12 ,28 ,54 ,96 ,142 ,242 2 integrated test 0 . 7 200 4 8 ,30 ,69 ,136 3 environmental test 2 260 6 2 ,21 ,52 , 96 ,148 ,212 4 1 . 2 300 3 10 ,80 ,200 high temperature burn2in 5 matching test 1 . 5 120 1 20 Ta ble 2 Converted test data of the space ca bin2borne electronic equipment ( )No . test item environmental factor ( )accumulated failures accumulated time to failure h begin2end time h 1 performance test 0 . 5 6 ,14 ,27 ,48 ,71 ,121 ,180,600 2 integrated test 0 . 7 185 . 6 ,201 ,228 . 3 ,275 . 2 180,3206,10 3 environment test 2 324 ,362 ,424 ,512 ,616 ,744 320,84010,16 4 840,120016,19852 ,936 ,1080 1 . 2 high temperature burn2in 5 matching test 1 . 5 1200,1380 19,20 1230 test during the detailed design phase . The continued. ECF of each test , the time of each test and fail2 Point estimates f or para meters of the ure distribution were given in table 1 . The re2 model Inserting the data into equations b = M sults of the conversions were given in table 2. t sb( ) , we obtain aM - 1/ ?ln and a = M/ t s Algorithm f or relia bility gro wth esti2 i = 1 t imation The algorithm begins with test data = 0 . 5222 , b = 0 . 5042 . This indicates that the preparedness and conversion. Then reliability corrective action is correct . Reliability of the can be estimated using the Crow model . A com2 product is growing. puter program for reliability growth analysis and 2 f it test Goodness of Using equation CM estimation of the equipment in development was 2 b M 6 t 1 i2 i - 1developed. ? = + ,we can obtain - i = 1 12 M 2 Mt s 2 2 α C= 0. 0385. From table , when = 0. 10 ,n Application to Equipment in 2 2 2 = 0 . 172 . Comparing with C , we C C ααn ,n , n Development f or Manned see that the data is related to the Crow model . Spaceflight Using the e2 Intervals estimate f or b 2 ()2bX M (γ) 1 - / 2 To assure a newly designed equipment for quation b= and b = L u ( ) manned spaceflight to meet the reliability re2 2 M - 1 2 quirement , it is important to dynamically moni2 ()bX 2 M (γ) 1 + / 2 ,we can obtain b= 0. 3576 , L tor the reliability growth fully using test data ( ) 2 M - 1such as performance test , environmental stress γb= 0. 7261 when confidence = 0. 90. U test , life test and matching test , etc . The meth2 Relia bility estimate and prediction Us2 ods described in this paper for a manned space2 1 - b t s - 1 flight cabin2mounted electronic equipment is a λ( )θ( ) t ing the equation t= =, s s a b practical example . The prototype test data are we can obtain the instantaneous MTBF is 136 .given in table 1 . Here , we made an assumption θ 85 h when T = 1380. Using the equation = L that every accident remedy worked immediately πθ( ) γand perfectly. The following procedure enables tand confidence ′= 0. 70 , the lower 1s a result of reliability growth analysis and estima2 confidence limit of MTBF is 108. 60 h. We can tion. also predict that MTBF is 199. 6 h when T = Test data conversion By using equation 1380 h + 1500 h. The curve of growth is shown () ( ) 2and equation 3the results of test data in figure 2. conversion are presented in table 2. Conclusion Gro wth trend test Inserting data of table 2 into the equation M The technique , which is composed of the U = , ( ?t- Mt/ tm/ 12 ( ) i ssCrow model and test data conversion TDC i = 1 method ,can dynamically monitor and track the we obtain U = - 10. 857. U= 1. 28 from α1 - / 2 reliability of the equipment and make full use of 2 αtablewhen = 0. 20 . Since U < - U, α1 - / 2 various test information in its development . It is reliability is increasing , analysis should be an economical and effective method and is worth to be popularized. [ Ref ere nce ] 1 Ushakov IA. Handbook of Reliability Engineering M . New York :John Wiley & Sons ,1994 :429,455HE Guowei ,DAI Cizhuang. Reliability Test Technology M . Bei 2 2 jing : National Defence Industry Press ,1995 :161,165( ) IEC TC - 56 CO150 . Reliability growth models and estimation 3 methods S. 4 ZHOU Yuanquan , WENG Chaoxi . Reliability Growth M . Bei 2 jing : Science Press ,1992 :64,75 GJB/ Z77 . A handbook of reliability growth management S . Bei 2 5 jing : The Center of Military Standard ,1995 :28,38 6 CHEN Jindun , SUN Hailong. A reliability growth assessment method and its application in the development of equipment in 120 7 space cabinJ . Space Medicine &Medical Engnieering ,1999 ,12 Meeker WQ , Hamada M. Statistical tools for the rapid development () 2:116,120 and evaluation of high2reliability products J . IEEE Trans on () 陈金盾 ,孙海龙 1 可靠性增长评估方法及其在航天产品研Rel , 1995 ,4 2:187,198 ( ) 制中的应用J . 航天医学与医学工程 ,1999 ,12 2: 116 , 书讯 响以及在特定环境中的生存技能 。该书适合于广大部队人类已大踏步迈入二十一世纪 , 新世纪带给我们的 是更加斑斓多姿的生活 。我们喜欢在网际遨游 ,也喜欢 ,医务人员 ,海上 官兵 、特种兵 、海军陆战队等部队的官兵 背上行囊去周游世界 。我们可以驾乘飞机环游地球 ,也 作业人员 ,院校师生 ,旅游探险者及其他野外工作者等 。可以乘风破浪远航 ,还可以驾驶汽车长途跋涉 ,更有趣的 当年轻的父母带领孩子去接受大自然的熏陶时 , 该书也 不失为一本开阔视野 、掌握生存本领的实用参考资料 。 是骑车或徒步旅行我们会被大自然的旖旎风光所吸 《遇险生存与营救》一书共分八章 ,约 30 万字 ,大 32 ,也有可能被无穷的诱惑带入险境 。当我们遭遇险情引 开本 ,暂定为每册 26 元 。由第四军医大学出版社近期出 时 ,我们不会束手无策 ,因为我们有挑战极限 、战胜自然 版 。 和自我的知识 、能力和魄力 。这种自豪源自一本书 ——— 征订联系方法 : 第四军医大学航空航天医学系 《遇险生存与营救》。 《遇险生存与营救》由航空救生专家孙喜庆教授主 姚永杰 邮编 710032编 ,图文并茂地论述了人在遭遇各种险境后求得生存和 :029 - 3374811 电话 组织营救的基本知识 ,重点阐述了海上 、热带丛林 、海港 、 传真 :029 - 3374381 寒区 、高原以及各种灾害条件下的环境特点 、对机体的影 E - mail : Kyxhd @ fmmu. edu. cn