节点文献

民航发动机健康管理中的寿命预测与维修决策方法研究

Methods of Life Prediction and Mainteance Decision Making in Civil Aeroengine Health Management

【作者】 戎翔

【导师】 左洪福;

【作者基本信息】 南京航空航天大学 , 载运工具运用工程, 2008, 博士

【摘要】 合理的民航发动机运营、维修与保障是提高其安全性、可靠性及经济性的重要手段,健康管理就是减少维修保障费用的一种重要方法,现在正受到越来越多的关注。本文针对民航发动机的运营与维修的特点,阐述了民航发动机健康管理的概念,对涉及的若干关键方法进行了深入研究,并开发了相应的集成维修管理系统,主要研究内容如下:(1)整体上,充分研究了国外几大典型的健康管理系统,在此基础上,从民航发动机运营与维修的角度出发,形成民航发动机健康管理的概念,设计了民航发动机健康管理系统的主要功能,阐述了民航发动机健康管理适用的关键方法,再基于开放式系统架构标准建立了应用技术体系。从方法到技术,再到应用,全面地构建了民航发动机健康管理方法与技术体系。(2)研究了民航发动机健康管理的关键方法之一:民航发动机在翼寿命预测与控制方法。首先阐述了故障强度概念和两种乘积型强度模型:比例危险模型和比例瞬时平均强度模型,研究了维修次数影响下的发动机使用可靠性建模方法。然后详细研究了两种在翼寿命预测方法:无状态监测参数时的寿命预测和有状态监测参数时的寿命预测。前者是先基于可靠性统计方法,由历史拆换记录进行威布尔建模,然后根据当前运行时间预测条件平均剩余寿命,进而得到在翼寿命的预测值,这是一种基于经验的预测;后者是一种混合模型,研究状态与发动机性能劣化衰退之间的联系,由比例危险模型建立状态与可靠性之间的关系,先得到发动机拆换控制限,结合状态变化趋势,预测发动机由于性能衰退引起的在翼寿命。再结合一些制造商、适航方面的时间限制,综合实现在翼寿命的控制。(3)研究了民航发动机健康管理的关键方法之二:民航发动机基于状态的维修决策方法。首先在现有各种概念基础上统一了视情维修与基于状态的维修,研究了视情维修决策要素与建模理论。深入研究了两种维修决策方法,一种是基于条件剩余寿命和维修前的条件概率,动态地调整传统的定时维修策略。另外一种决策方法是基于比例危险模型,可以根据发动机当前状态,以故障强度为维修阈值决策发动机的拆换和送修时机,或以单位时间期望总维修费用最小标准下决策最优的预防性维修间隔等。(4)针对目前监测软件与维修管理软件相互孤立的问题,并考虑航空公司的实际需求,基于民航发动机健康管理的技术体系,设计了民航发动机集成维修管理系统的业务流程与主要功能模块,使用Oracle数据库构建了系统的后台数据库,采用Java进行系统的应用程序开发,目的是实现民航发动机状态监测、寿命控制、维修决策与资源调度。此系统也是上述健康管理方法与技术的验证,正逐步应用于航空公司的发动机日常工程管理中,应用效果非常显著,提高了工作效率,降低了维修成本。

【Abstract】 The appropriate civil aeroengine operation, maintenance and support are important for high safety, reliability and economics. Engine health management is an important method to decrease maintenance and support costs. And researchers are paying much attention to this method gradually. In this thesis, according to the features of civil aeroengine operation and maintenance, the concept of civil aeroengine health management is introduced, some involved key methods are studied in detail, and an integrated aeroengine maintenance management system is developed. The main research content is as following.(1) Holisticly, based on some famous overseas health management systems, the concept of civil aeroengine health management is introduced for engine operaton and maintenance, and main functions of this health management system are designed. After studying the involved key technologies and application system architecture based on the standards of open system architechture. So the general civil aeroengine health management method and system are established, from method to technologies, and to applications.(2) One key method of civil aeroengine health management is life on wing (LOW) prediction and control. Firstly the concept of failure intensity and two multiplative intensity models, which are proportional hazards model and proportion mean intensity model, are introduced. A specified engine operation reliability model with effect of repair numbers is studied. Then two kinds of LOWprediction models are studied detailedly. The first model is LOW prediction only using time data without condition monitoring data. This model is also called experience-based prognostics, which is a kind of statistics reliability model. Based on the failure time historyical data, the Weibull model is established and the conditional mean residual life is calculated at present time. So the LOW can be predicted. The second LOW prediction model is a kind of hybrid model which study the connection between engine conditions and performance deterioration. Using proportional hazards model (PH model) which combine the condition and reliability data, the LOW due to performance deterioration can be predicted with the engine removal control limit and condition trend. Then LOW control can be completed with the time constrains from engine manufacturer and airworthiness directives.(3) In this dissertation the other key method of civil aeroengine health management is condition based maintenance (CBM) decision-making. The relationship beween CBM and on condition maintenance is consolidated. The factors and theory of CBM decision-making model are explained. Two models are studied deeply. The first one is a dynamicly adjusted model according to the conditional residual life and conditional probability. Also based on the PH model, according to the present status, civil aeroengine removal and shopvisit can be decided with the threshold of failure intensity. And the optimal engine preventive maintenance interval is offered in term of minimal expected total maintenance cost per unit time.(4) Considering the practical demands of airlines and the fact of isolation between engine condition monitoring softwares and maintenance management softwares, based on the civial aeroengine health management system architecture, an integrated aeroengine maintenance management system is designed including the work flows, main modular functions and database. This system can implement engine condition monitoring, LOW control, maintenance decision-making and optimal engine schedule. The system is developed using Java programs and Oracle database software. And this integerated sytem is being applied to the aeroengine engineering management in airlines effectively. The work efficiency is increased and the maintenance cost is decreased.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络