基于Bayes的导弹加速贮存试验参数估计

Estimation of Parameters for Missile Accelerated Storage Life Tests Based on Bayes Theory

目的研究导弹加速贮存寿命试验中的参数估计问题。方法拟在小子样的情况下,设计一系列反复连续的导弹加速贮存仿真试验,将前一试验的结果作为之后试验的先验分布,基于Bayes方法建立导弹加速贮存寿命模型,给出贮存寿命参数估计算法,探讨导弹贮存可靠性评估方法。结果算例给出贮存寿命参数的极大似然估计,利用RBA方法对参数估计值进行降偏修正,得出参数的修正似然估计值,给出导弹贮存可靠度。给出的参数估计在均方误差意义下结果较优。结论序列试验Bayes算法易行、精度高,可有效对导弹贮存可靠性进行评估。

Objective To study the question of parameter estimation for missile accelerated storage life tests. MethodsA series of repeated and continuous accelerated simulation tests for missile storage life were designed with insufficient data. The results of previous test was taken as prior distribution of the next test. An accelerated storage life model of missile was established based on Bayes method to propose the algorithm of estimation of storage life parameters and discuss methods for reliability assessment of missile storage. Results The example provided the maximum likelihood estimation of parameters. The value of estimation was adjusted with RBA method and the storage reliability of missile was proposed. The mean square errors of the new estimation are less. Conclusion Bayes algorithm of sequence tests is simple and precise, which will be applicable to evaluation of missile storage reliability.