基于GA的可靠性寿命分布参数的极大似然优化估计  

Maximum Likelihood Optimization Estimation for Parameters of Reliability Life Distribution by GA

作　　者：方华元[1] 胡昌华[1] 樊红东[1] 丁力[1] 

机构地区：[1]西安第二炮兵工程学院,陕西西安710025

出　　处：《上海航天》2006年第2期50-53,共4页Aerospace Shanghai

摘　　要：基于极大似然估计法的基本原理和优化理论,建立了可靠性寿命分布参数的极大似然优化估计模型,并采用遗传算法(GA)进行求解,分析了GA在模型求解中的编码、选择策略、交叉和变异等关键步骤,以及约束条件的处理。给出了用该法估计单威布尔分布和混合威布尔分布参数的一实例。结果表明,工程上此法可有效估计可靠性寿命分布的参数,所获结果更为精确合理。An optimization model for maximum likelihood parameters estimation of reliability life distribution was established based on the principle of the maximum likelihood functions method and theories of optimization in this paper. The model was solved by genetic algorithm, and the key steps, such as code, selection, cross＇ and mutation, and the method of disposing the problem of restriction were given. The processing for the distribution parameters of Weibull distribution and mixed Weibull distribution for some example was presented. The result showed that the＇devdoped method would be to estimate the parameters of reliability life distribution in engineering availably, the restdt obtained wotrld be more accuracy and in reason.

关 键 词：可靠性寿命 遗传算法 极大似然估计 威布尔分布 优化 参数估计 

分 类 号：TB114.3[理学—概率论与数理统计]