Application Of Small Circumstance In Weapon System Accuracy Assessment And Reliability Growth Analysis

According to the dominant characteristics and necessary of test analysis and assessment of weapon system and reliability growth analysis in small sample circumstance, the Bayesian method in small sample statistical inference and dynamic statistical theory is adopted as an important and suitable approach. Some general theoretical problem in Bayesian approach's engineering applications have been discussed and concluded in this dissertation, and what is more, the dissertation conducted further application researches in respect of windage correcting weapon system accuracy and dynamic reliability growth assessment.First, as the base of application research, some general theoretical problems in Bayesian method and prior information processing are introduced in this dissertation. On the basis of engineering application, the accuracy assessment of WCMD (windage correcting machine of delivery) is completed in small sample circumstance: based on the special character of destroying mechanism and attacking mode of the WCMD, the scheme of accuracy assessment for WCMD is especially presented by assessing the destroying firepower intensity, the destroying effect and the uniformity of delivery. An improved Bayes SPRT method is presented for expediting the speed of convergence and reducing the times of test, furtherly, the method of assessing the uniformity of delivery is studied. A simulated demonstration is given.Secondly, as the fore work of the dynamic statistical fund item, the Bayesian method is adopted in the assessment of reliability growth analysis. Based on characteristics of the reliability growth testing of the complex system which possesses the trait of different matrix, small sample circumstance and excellent capability, combined with linear model theory, an improved method of Bayesian iterative assessment to changeable parameters in the multilayer prior information condition is presented for enhancing the capability of error eliminating. A simulated demonstration is given.At last, based on characteristics of the reliability growth, the maneuverability of AMSAA model in small simple circumstance is studied. Respecting to the practical signification of Duane model, the linear model of invalidation probability is present for being as an important and suitable approach to the iterative model. The multistage Bayesian assessment for AMSAA model and a simulated demonstration are given in this dissertation.The insufficient of this dissertation is that: the scheme of WCMD hasn't been used in practice because the dismatching the practical research and progress. This is a pity of the mission.