A small-sample evaluation method for the reliability of explosive products with high confidence

Abstract

The invention discloses a high-confidence small-sample evaluation method for the reliability of explosives. According to the mensurability and reliability requirements of explosives, and under the premise that the performance parameters of explosives obey normal distribution or approximate normal distribution or the evaluated performance parameters after transformation obey normal distribution or approximate normal distribution, the parameters and uncertainty of a population distribution model are estimated based on small sample data and measurement uncertainty of product performance parameter quantitative measurement. In the absence of priori information and in the presence of statistical characteristics, the number of pieces of sample data needed by the method can be reduced to five under the condition that the confidence level is not less than 0.90 and the reliability of explosive performance evaluation is not less than 0.99 or above. The confidence level of evaluation results reaching required reliability is not lower than 97.5% of the prescribed confidence level.

Description

technical field [0001] The invention relates to a quantifiable high-value explosive product performance reliability evaluation technology, in particular to a high-confidence small-sample evaluation method for the reliability of explosive products. Background technique [0002] In the design, development, test and batch production of products with high reliability requirements, how to obtain the performance data required by the product through a small amount of reliability test and evaluate whether it meets the requirements is a hot spot in the field of reliability research today. . However, so far, various classic small-sample evaluation methods proposed by research at home and abroad either still require large sample size data to ensure that the accuracy and credibility of the evaluation results are within an acceptable range, or although the evaluation The required sample size data has been reduced, but the precision and reliability of the assessment results have been que...

AI Technical Summary

Problems solved by technology

Other new evaluation methods that are being researched and developed and may be applied in the future, such as fuzzy reliability evaluation method, QMU (quantification of margin and its uncertainty) evaluation method, although each has certain advantages, they are still subjective in the operation process. There is arbitrariness in quantification or there are difficulties in quantification objectively, so there is still a certain time distance before these evaluation methods are applied and produce benefits

Images