基于AMSAA模型的系统可靠性评估方法研究

针对航空机载设备可靠性增长试验数据不规则的特点，在AMSAA模型的基础上提出了识别异常点的AMSAA模型与处理截断数据的AMSAA模型，并给出了模型拟合优度检验方法。经实例验证表明，识别异常点的AMSAA模型可以在给定置信度下识别出异常点并排除异常点对瞬时MTBF（平均故障间隔时间）极大似然估计值的干扰；截断数据的AMSAA模型能够利用被截取的部分数据，得到准确的瞬时极大似然估计值。

Research on Estimation of System Reliability Based on AMSAA Model

To Research the irregular data of aero equipment reliability growth test,tow models were advanced based on the theory of AMSAA model.The one model is to identify exceptional points,and the other one is to deal with truncated data.The goodness of fit test for the tow models were also presented.According to the example analysis,the first model can identify exceptional points in the condition of given belief and get rid of the influence to the maximum likelihood estimation of instantaneous MTBF.The other model can get the maximum likelihood estimation of instantaneous MTBF exactly by the truncated data.