航空兵器
航空兵器
항공병기
HANGKONG BINGQI
2014年
6期
59-61
,共3页
成败型产品%混合Beta先验分布%继承因子%可靠性
成敗型產品%混閤Beta先驗分佈%繼承因子%可靠性
성패형산품%혼합Beta선험분포%계승인자%가고성
success/failure product%mixed Beta prior distribution%inheritance factor%reliability
成败型产品可靠性的小样本评估方法,把历史试验数据作为Bayes估计方法的先验信息,有效地减少了样本量。但是传统的Bayes方法过度依赖验前信息,当新产品有改进时造成鉴定结果的冒进。针对过度依赖先验信息问题,本文提出了使用混合Beta先验分布,引入继承因子,通过调整继承因子的大小来控制对验前信息的依赖程度,考虑了成败型产品的变化对评估结果的影响。采用该方法对某导弹的可靠性进行了评估。
成敗型產品可靠性的小樣本評估方法,把歷史試驗數據作為Bayes估計方法的先驗信息,有效地減少瞭樣本量。但是傳統的Bayes方法過度依賴驗前信息,噹新產品有改進時造成鑒定結果的冒進。針對過度依賴先驗信息問題,本文提齣瞭使用混閤Beta先驗分佈,引入繼承因子,通過調整繼承因子的大小來控製對驗前信息的依賴程度,攷慮瞭成敗型產品的變化對評估結果的影響。採用該方法對某導彈的可靠性進行瞭評估。
성패형산품가고성적소양본평고방법,파역사시험수거작위Bayes고계방법적선험신식,유효지감소료양본량。단시전통적Bayes방법과도의뢰험전신식,당신산품유개진시조성감정결과적모진。침대과도의뢰선험신식문제,본문제출료사용혼합Beta선험분포,인입계승인자,통과조정계승인자적대소래공제대험전신식적의뢰정도,고필료성패형산품적변화대평고결과적영향。채용해방법대모도탄적가고성진행료평고。
Evaluation method for small sample of success /failure product reliability takes historical test data as prior information of Bayes, effectively reducing the quantity of samples.But over-reliance on prior information in the traditional Bayes method, will get the aggressive identification results when new products are improved.To deal with the problem of over-reliance on prior information, this paper puts for-ward using Bages evaluation based on mixed Beta prior distribution, which introducing inheritance factor. Thus the dependence on prior information is controlled by adjusting the inheritance factor, and the effect of changes of success/failure product on evaluation result is considered.With this method, the missile re-liability is assessed.