基于Bayes方法的民机冗余部件随机可靠性评估  

Random Reliability Evaluation of Components for Civil Aircrafts Using Bayes Method

作　　者：赵飞[1] 王华伟[1] 王岩峰[1] 

机构地区：[1]南京航空航天大学民航学院,南京210016

出　　处：《机械科学与技术》2011年第1期6-9,共4页Mechanical Science and Technology for Aerospace Engineering

基　　金：国家自然科学基金与民航总局联合项目(60879001);国家自然科学基金重点基金项目(60939003)资助;江苏省自然科学基金项目(SBK200922876)

摘　　要：由于民机部件的故障发生具有随机性的特点,笔者采用马尔可夫链描述系统的状态变化,通过拉普拉斯变换得到部件可靠性相关指标函数,同时鉴于民机部件可靠性信息较少,属于小样本评估,采用Bayes方法,充分融合多源信息,结合部件检测信息和历史信息等,通过最大熵原则,实现验前信息的转化与先验参数的求解,并对两发动机组成的冗余动力系统进行了可靠性评估。通过算例,验证了笔者提出方法的有效性。Failures of civil aircraft components occur with random characteristics.In this paper,by using markov chain to describe the system state changes,the relevant reliability index functions are obtained by Laplace transform.Considering that there is little reliability information of the components for civil aircraft which belongs to assessment of small sample,we use Bayes method and make full use of multi-source information,such as detection information and history information.By the principle of maximum entropy,we achieve the transformation of priori information and the solution of the priori parameters.We conducted reliability assessment of a redundant power system consisting of two engines.An example is presented in the end to verify the effectiveness of the method proposed in this paper.

关 键 词：民机 马尔可夫 BAYES 可靠性 多源信息 

分 类 号：O213[理学—概率论与数理统计]