民航飞机飞行管理系统可靠性监控仿真研究

摘要

Reliability monitoring of civil aviation flight management system can not only ensure the safety of the flight,but also avoid excessive maintenance.Reliability modeling is the base of the system reliability monitoring,but lacking sample data is the important limiting factor of the system reliability modeling.In order to improve the accuracy of system reliability modeling,a Bayesian reliability assessment method of small sample failure data for civil aviation aircraft flight management computer system with predictive maintenance was proposed.The selection of prior and posterior distributions for Weibull model was discussed.Combined with examples,Bayesian point and interval estimates of model parameters and reliability indices including mean time to faiiure and reliability etc.were obtained through Markov chain Monte Carlo method.The result shows that Bayesian reliability assessment method has a higher accuracy than maximum likelihood estimate in the reliability evaluation of civil aviation aircraft flight management computer system.The proposed reliability model based on Bayesian method is also applicable to the reliability analysis of other aircraft system components,whose failure distribution conforms to the Weibull model.%民航飞机飞行管理系统可靠性监控既可以保证飞机飞行安全,又可以避免过度维修.可靠性建模是实现系统可靠性监控的基础,而样本数据少是系统可靠性建模的重要限制因素.为了提高系统可靠性建模的准确度,提出了在少样本故障数据下,民航飞机飞行管理系统预防性维修的贝叶斯可靠性评估方法,分析了威布尔模型的先验分布和后验分布.结合实例,利用马尔科夫链蒙特卡洛方法获得了模型参数、平均寿命、可靠度等可靠性指标的贝叶斯点估计以及区间估计,结果表明,在民机飞行管理系统的可靠性分析中,与极大似然估计法相比,贝叶斯方法具有较高的精度.提出的基于贝叶斯方法的可靠性模型也适用于其它符合威布尔模型的飞机系统部件的可靠性分析.