SPOT和Bayes递推估计融合的运载火箭样本量设计
2025,47(1):207-213
摘要：
针对运载火箭小子样条件，结合序贯验后加权检验(sequential posterior odd test, SPOT)和Bayes递推估计法，分别从假设检验和参数估计两方面，对传统试验样本量评估方法进行改进。在对运载火箭服从正态分布的性能指标进行评估时，引入复合等效系数来有效融合多源数据，弥补真实试验数据或现场数据的不足。综合考虑两类风险和置信度要求，制定合理的评估方案，有效减少所需试验样本数量，从而控制试验成本。通过算例分析发现，提出样本量评估方法结果真实可信，能够有效降低样本量需求，可较好用于小子样条件下的运载火箭样本量试验设计。
针对运载火箭小子样条件，结合序贯验后加权检验(sequential posterior odd test, SPOT)和Bayes递推估计法，分别从假设检验和参数估计两方面，对传统试验样本量评估方法进行改进。在对运载火箭服从正态分布的性能指标进行评估时，引入复合等效系数来有效融合多源数据，弥补真实试验数据或现场数据的不足。综合考虑两类风险和置信度要求，制定合理的评估方案，有效减少所需试验样本数量，从而控制试验成本。通过算例分析发现，提出样本量评估方法结果真实可信，能够有效降低样本量需求，可较好用于小子样条件下的运载火箭样本量试验设计。
基金项目：
国家自然科学基金资助项目（62103422）；湖南省自然科学基金资助项目（2021JJ40680）
国家自然科学基金资助项目（62103422）；湖南省自然科学基金资助项目（2021JJ40680）
Sample size design of launch vehicle combined with SPOT and Bayes recursive estimation
HUANG Pengqizi
College of Sciences, National University of Defense Technology, Changsha 410073 , China
DUAN Xiaojun
College of Sciences, National University of Defense Technology, Changsha 410073 , China
ZHANG Yinhui
Beijing Institute of Tracking and Telecommunications Technology, Beijing 100094 , China
College of Sciences, National University of Defense Technology, Changsha 410073 , China
DUAN Xiaojun
College of Sciences, National University of Defense Technology, Changsha 410073 , China
ZHANG Yinhui
Beijing Institute of Tracking and Telecommunications Technology, Beijing 100094 , China
Abstract：
Aiming at the small sample data of launch vehicle, combining the SPOT(sequential posterior odd test）and Bayes recursive estimation method to improve the traditional sample size design from hypothesis testing and parameter estimation, respectively. When evaluating the performance indexes of launch vehicle following the normal distribution, the composite equivalent coefficient was introduced to fuse multi-source data, which can make up for the shortage of experimental data or test data. Considering the two types of risks and confidence requirements comprehensively, a reasonable evaluation scheme was designed. Through the analysis of examples, the results verified that the proposed method are authentic and credible, which can reduce the sample size demand effectively, and can be used for the sample size test design of launch vehicle under the small sample condition.
Aiming at the small sample data of launch vehicle, combining the SPOT(sequential posterior odd test）and Bayes recursive estimation method to improve the traditional sample size design from hypothesis testing and parameter estimation, respectively. When evaluating the performance indexes of launch vehicle following the normal distribution, the composite equivalent coefficient was introduced to fuse multi-source data, which can make up for the shortage of experimental data or test data. Considering the two types of risks and confidence requirements comprehensively, a reasonable evaluation scheme was designed. Through the analysis of examples, the results verified that the proposed method are authentic and credible, which can reduce the sample size demand effectively, and can be used for the sample size test design of launch vehicle under the small sample condition.
收稿日期：
2022-09-16
2022-09-16
