新型弹药可靠性评估方法

Reliability Evaluation Method of New Ammunition

为了研究新型弹药可靠性评估方法,采用混合贝塔分布为先验分布的贝叶斯方法对某新型弹药进行可靠性评估。分析贝叶斯方法的适用条件,并将之推广到多先验情况。通过实例比较该方法与经典方法和使用贝塔分布为先验分布的贝叶斯方法的差异。结果表明:该方法能更为合理地用于新型弹药可靠度评定,充分利用产品先验信息,有效降低现场试验所需的样本量,很好地避免由于小样本数据被大量先验信息"吃掉"而导致评定结果的"冒进"现象。

For researching new ammunition reliability evaluation method, use mixed Beta distribution as prior distribution Bayesian method to carry out reliability evaluation method for new ammunition. Analyze the application condition of Bayesian method, and use it to multi-prior condition. Compare the method with the classical method, and Bayesian method with Beta distribution by example. The results show that the method can be more effectively in new ammunition reliability evaluation, make full use of product prior information, reduces sample quantity in test. Therefore avoid small sample data are eaten large quantity of prior information and cause ＂premature advance＂ of evaluation.