作者：王京¹, 何华锋¹, 李天梅¹, 徐从启², 杨宗浩¹, 韩聪聪¹

作者单位：1. 火箭军工程大学控制工程系, 陕西 西安 710025;
2. 解放军总后勤部建筑工程技术研究所, 陕西 西安 710032

关键词：测试性评估;相容性检验;增长试验;Fisher统计量

摘要：

在测试性评估过程中，为检验测试性增长试验数据与现场数据的总体一致性，提出一种基于Fisher统计量的相容性检验方法。主要针对小子样、成败型数据，提出顺序约束模型完成对试验数据的增长趋势检验，依据超几何分布构造Fisher统计量，对相邻增长阶段间连带关系进行判定，继而得到增长试验数据与现场数据总体一致性的相关结论。实例应用表明：在小样本情况下所提方法能对总体一致性做出合理判定，其可处理的样本量上限为26，工程实际中，在给定显著性水平的情况下，应综合考虑增长成本、结果稳定性、维修改进技术水平等因素的影响，合理安排装备增长过程中的增长阶段数。

The method of compatibility check of testability growth test data based on Fisher-statistics

WANG Jing¹, HE Huafeng¹, LI Tianmei¹, XU Congqi², YANG Zonghao¹, HAN Congcong¹

1. Department of Control Engineering, Rocket Force University of Engineering, Xi'an 710025, China;
2. The Institute of Construction Engineering Research, General Logistics Department of PLA, Xi'an 710032, China

Abstract: The method of compatibility check based on Fisher-statistics is put forward to verify the consistency of testability growth test data and field data during the process of testability evaluation. Aiming at the data of small sample and success or failure, a sequential constraint model is proposed to test the growth trend of test data, constructing Fisher-statistics based on hypergeometric distribution and judging the relationship between two adjacent stages, then the conclusions about the overall consistency between the growth test data and the field data are obtained. Validation of the method is presented via application examples. Results indicate that the method can also make a reasonable decision about the overall consistency in small sample sizes whose number is less than 26. In practical engineering, it's a better choice to set growth stage after synthesizing the influence of cost, stability and technical level at a given level of significance.

Keywords: testability evaluation;compatibility check;growth test;Fisher-statistics

2017, 43(8): 11-13,40  收稿日期: 2016-09-22;收到修改稿日期: 2016-11-18

基金项目: 国家自然科学基金项目（61304103）

作者简介: 王京(1992-),男,河南郑州市人,硕士研究生,专业方向为测试性验证与评估。

参考文献

[1] 中国航空工业总公司三〇一所.装备测试性大纲:GJB 2547-1995[S]. 1996.
[2] 田仲,石君友. 系统测试性设计分析与验证[M]. 北京:北京航空航天大学出版社,2003:173-182.
[3] 张伟昆. 测试性分析与评估体系的研究[J]. 国外电子测量技术,2015,34(5):38-42.
[4] 常春贺,杨江平,曹鹏举. 基于研制信息的测试性验证试验方案研究[J]. 航空学报,2012,33(11):2057-2064.
[5] 徐萍,刘松林,李勇. 测试性试验概念及模型研究[J]. 计算机测量与控制,2006,14(9):1149-1152.
[6] 李天梅. 装备测试性验证试验优化设计与综合评估方法研究[D]. 长沙:国防科学技术大学,2010.
[7] 刘晗,郭波. 小子样产品可靠性Bayes评定中的相容性检验方法研究[J]. 机械设计与制造,2007(5):165-166.
[8] LI T M, XU C Q, QIU J, et al. The assessment and foundation of bell-shaped testability growth effort functions dependent system testability growth models based on NHPP[J]. Mathematical Problems in Engineering,2015,2015:1-17.
[9] 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(6):31-36.
[10] 张湘平,曹国敏. 验前信息与现场子样的相容性检验方法研究[J]. 飞行器测控学报,2002,21(1):55-59.
[11] 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.
[12] 胡正东,曹渊,张士峰,等. 特小子样试验下导弹精度评定的Bootstrap方法[J]. 系统工程与电子技术,2008,30(8):1493-1497.
[13] 茆诗松. 统计手册[M]. 北京:科学出版社,2003:177-185.
[14] LOVE J J. Fisher statistics[C]//Encyclopedia of Geomagnetism and Paleomagnetism Springer Netherlands,2007.
[15] 李明明. Fisher与小样本及"n维几何法"[D]. 天津:天津财经大学,2014.
[16] LIN T C. Development of U.S. air force international ballistic missle weapon systems[J]. Journal of Spacecraft and Rockets,2003,40(4):491-509.