改进多源信息加权融合的验前分布确定方法研究

Research on the Improved Method of Determining the Pre-test Distribution of Multi-source Information Weighted Fusion

为给出比较准确的验前分布,提高装备维修性定量指标验证结果的可信度,提出了一种改进多源信息加权融合的验前分布确定方法。结合Bootstrap法,改进了子样本重抽样方式,拓展了子样本数据的取值范围,确定了各验前信息的验前分布。改进了基于质量因子的融合权重确定方法,定义了均值可信度权重因子和样本量可信度权重因子,综合两种权重因子,使所求权重更加科学合理。结合某型装甲装备维修性验证试验,对MTTR指标进行分析,结果表明所提方法融合后的验前分布更加准确,更接近真实分布,对提高装备维修性验证结果的可信度具有重要的参考价值。

In order to give a more accurate pre-test distribution and improve the reliability of the verification results of equipment serviceability quantitative indicators,a method was proposed for the improvement of determining the pre-test distribution of multi-source information weighted fusion.In combination with the Bootstrap method,the sub-sample resampling method was improved,the value range of sub-sample data expanded,and the pre-test distribution of each pre-test information determined.The method for determining the fusion weighting method based on quality factor was improved,mean reliability weighting factor and sample size reliability weighting factor were defined,and the two weighting factors were integrated to make the target weights more scientific and reasonable.The MTTR index was analyzed in combination with the serviceability verification test of a certain type of armor equipment.The results show that the pre-test distribution of the proposed method is more accurate and closer to the real distribution after the fusion of the proposed method,which has important referential value for improving the reliability of the equipment serviceability verification results.