摘要
为满足军品验收时,既能保证样本容量小,又能保证质量检验标准不变的要求,采用贝叶斯方法对某型通用装备的验收费用控制进行了优化仿真。首先收集并处理所得到的验前信息,得出先验分布,其次根据所求得的验前超参数求出后验分布,最后使用MATLAB软件进行仿真验证,求得区间分布进而得出所需要的样本容量,并与军品质量检验验收普遍采用的经典算法进行对比分析,并最终得出可靠的结论。结果显示,在军品验收费用控制问题上,从控制样本量的角度使用贝叶斯方法对军品验收普遍采用的经典算法进行优化改进,能够达到既能保证抽样标准不变,又能在减少样本容量方面达到节省军品验收费用的目的。具有较强的工程应用价值。
To ensure the small sample size and high quality inspection standard, this paper uses Bayes Methods to optimize the acceptable cost control of a certain type of general-purpose equipment. Firstly, the prior distribution is obtained by collected and handled prior information, then the posterior distribution is derived according the obtained prior hyper-parameters, and the zone distribution is obtained through MATLAB simulation to find the simple size and compared with the result of classic algorithms which has been widely used in material acceptance. Finally a reliable conclusion is drawed. The result shows that the proposed method can improve the classic algorithms. It can not only keep up the acceptable level, but also can save the military products acceptance cost in reduce the sample size.
出处
《计算机仿真》
CSCD
北大核心
2015年第9期34-38,共5页
Computer Simulation
基金
国家自然科学基金(J0910014)
关键词
军品验收
样本容量
贝叶斯方法
验收费用
Military products acceptance
Sample size
Bayesian methods
Acceptance cost