在航空航天领域由于成本、时间周期等原因进行疲劳寿命及可靠性评估时样本量通常极少（m=1或2），利用相容性检验方法可对样本量进行扩充。常规的Wilcoxon秩和检验和K-S（Kolmogorov-Smirnov）检验适用于小样本情形，而极小样本相容性检验方面研究较少，且缺乏对方法合理性的详细说明和对不同方法检验功效优劣的比较。航空航天产品疲劳寿命多服从正态分布，因此本文主要以正态分布作为研究对象。利用Monte Carlo仿真发现从某一正态分布N（μ，σ2）中随机抽取两个样本x1、x2计算均值μ1和标准差σ1后构建新正态分布N（μ1，σ12），其±σ1、±2σ1和±3σ1范围内的点落在原正态分布N（μ，σ2）±3σ范围内的概率依次为99.80%、98.13%和97.37%。在此基础上针对现场试验数据样本量为2的情况，本文提出利用3σ原则对先验信息数据进行相容性检验从而扩充样本量的方法。将该方法与两种文献方法对比后发现其误差率明显更低并呈现出检验性能随先验数据增加而不断提高的优势。
When evaluating the fatigue life and reliability in the aerospace field, the locale test data sample size is usually extremely small (m=1 or 2) due to the cost and time limits. The compatibility test method can be used to expand the sample size. Conventional Wilcoxon rank sum test method and K-S (Kolmogorov-Smirnov) method are applicable for situation of small sample size. Less research has been conducted on the method for compatibility test of minimum sample size, and there is a lack of detailed explanation of the rationality of the method and comparison of actual effects of different methods. The fatigue life of aerospace products usually obeys the normal distribution, so normal distribution is analyzed in this paper. Two points x1,x2 are randomly selected from a normal distribution N(μ,σ2), and the mean μ1 and standard deviation σ1 are calculated to construct the new normal distribution N(μ1,σ12). It is found using the Monte Carlo simulation that the probabilities that points placing at ±σ1, ±2σ1 and ±3σ1 ranges of the new normal distribution N(μ1,σ12) place at ±3σ range of the original normal distribution N(μ,σ2) are 99.80%,98.13% and 97.39% respectively. Aiming at the situation that the population follows normal distribution and the sample size is 2, this paper proposes to use the 3σ principle to test the prior information data and thus to expand the sample size. A comparison with other two methods shows that with the proposed method, the error rate is obviously lowered, and with the increase of prior information data, the method performs better.
[1] 周兆麟. 数理统计学[M]. 北京:中国统计出版社, 1987:4. ZHOU Z L. Mathematical statistics[M]. Beijing:National Bureau of Statistics of China, 1987:4(in Chinese).
[2] 施雨. 应用数理统计[M]. 西安:西安交通大学出版社, 2006:4. SHI Y. Introduction of applied statistics[M]. Xi'an:Xi'an Jiaotong University Press, 2006:4(in Chinese).
[3] KIJEWSKI T, KAREEM A. Bayes estimation of a langevin mean direction[J]. Communications in Statistics-Simulation and Computation, 2017, 46(4):2769-2783.
[4] FABRICE G, BERNARD D, EMMANUEL U. Reliability estimation by Bayesian methods:Definition of prior distribution from architecture of the system[J]. Reliability Engineering and System Safety, 2013, 82:299-306.
[5] SHONOSUKE S. On bootstrap averaging empirical Bayes estimators[J]. Statistics, 2017, 55(2):331-357.
[6] LÓPEZ J P, KRZYWINSKI M, ALTMAN N. Points of significance:Bayes' theorem[J]. Nature Methods. 2015, 12(4):277-278.
[7] LI H Y, ZUO H F, SU Y, et al. Study on segmented distribution for reliability evaluation[J]. Chinese Journal of Aeronautics, 2017, 30(1):310-329.
[8] 罗强, 骆飞. 基于Bootstrap法的小样本岩土参数标准值计算[J]. 华中科技大学学报, 2017, 45(8):93-98. LUO Q, LUO F. Calculating formula for standard value of small-sample geotechnical parameters based on Boot-strap method[J]. Journal of Huazhong University of Science and Technology, 2017, 45(8):93-98(in Chinese).
[9] 郭荣化, 李勇. 小样本试验中验前数据与现场数据相容性检验方法研究[C]//第十三届中国系统仿真技术及其应用学术会议论文集. 北京:中国自动化学会系统仿真专业委员会, 2011:1115-1121. GUO R H, LI Y. Research on checkout methods for the compatibility test between the pre-test data and the local-test data in small sample test[C]//Proceedings of 13th Chinese Conference on System Simulation Technology & Application (CCSSTA2011). Beijing:Chinese Association of Automation and System Simulation Committee, 2011:1115-1121(in Chinese).
[10] 蔡洪, 张士峰, 张金槐. Bayes试验分析与评估[M]. 长沙:国防科技大学出版社, 2004:73-74. CAI H, ZHANG S F, ZHANG J H. Bayes test analysis and evaluation[M]. Changsha:NUDT Press, 2004:73-74(in Chinese).
[11] 陈涛. 小样本条件下风电齿轮传动系统动态可靠性预测方法[D]. 大连:大连理工大学, 2012:40-56. CHEN T. Prediction methods of dynamic reliability for wind gear transmission under the conditions of small samples[D]. Dalian:Dalian University of Technology, 2012:40-56(in Chinese).
[12] 刘冰月, 王景芹. 低压成套开关设备的先验信息相容性检验技术[J]. 东北电力大学报, 2013, 33(1-2):14-19. LIU B Y, WANG J Q. Compatibility check of the prior information on low voltage switchgear[J]. Journal of Northeast China Institute of Electric Power Engineering, 2013, 33(1-2):14-19(in Chinese).
[13] 王江元. 验前试验数据与现场试验数据相容性检验方法[J]. 战术导弹技术, 2004(2):10-12. WANG J Y. Checkout methods for compatibility of pre-test data and locale test data[J]. Tactical Missile Technology, 2004(2):10-12(in Chinese).
[14] KESELMAN H J, OTHMAN A R, WILCOX R R. The new and improved two-sample test[J]. Psychological Science, 2004, 15:47-51.
[15] 李鹏波, 张士峰, 蔡洪. 导弹仿真试验结果静态一致性检验及其置信度[J]. 现代防御技术, 1999, 27(6):31-36. LI P B, ZHANG S F, CAI H. Checkout methods for compatibility of pre-test data and locale test data[J]. Modern Defence Technology, 1999, 27(6):31-36(in Chinese).
[16] 张金槐. 多源信息的Bayes融合精度鉴定方法[J]. 国防科技大学学报, 2001, 23(3):93-97. ZHANG J H. Bayes fusion precision identification method for multi-source information[J]. Journal of National University of Defense Technology, 2001, 23(3):93-97(in Chinese).
[17] MOSER B K, STEVENS G R, WATTS C L. The two-sample test versus Satterwaite's approximate F test[J]. Communications in Statistics Theory and Methods, 1989, 18(11):3963-3975.
[18] NEUHAUSER M. Two-sample tests when variances are unequal[J]. Animal Behaviour, 2002, 63(4):823-825.
[19] KASUYA E. Mann-Whitney U-test when variances are unequal[J]. Animal Behaviour, 2001, 61:1247-1249.
[20] 唐雪梅. 小样本场合下相容性检验方法[J]. 系统工程与电子技术, 2001, 23(10):66-68. TANG X M. Consistency test methods in small sample situation[J]. Systems Engineering and Electronics, 2001, 23(10):66-68(in Chinese).
[21] 王晖, 潘高田, 臧兴震. 正态分布的小样本数据的相容性检验理论和方法[J]. 数学的实践与认识, 2005, 35(10):131-137. WANG H, PAN G T, ZANG X Z. Theories and methods of compatibility test on normal distribution with few samples[J]. Mathematics in Practice and Theory, 2005, 35(10):131-137(in Chinese).
[22] 张湘平, 张金槐. 验前信息与现场字样的相容性检验方法研究[J]. 飞行器测控学报, 2002, 21(1):55-59 ZHANG X P, ZHANG J H. Study on consistency test methods of pretest information and field subsamples[J]. Journal of Spacecraft TT&C Technology, 2002, 21(1):55-59(in Chinese).
CJA