Aeroengines In A Fleet Oriented Maintenance Programmingand Its Key Technologies

The development of modern manufacturing service and change of product value chain show that the equipment maintenance management is becoming increasingly important. An aeroengine is the heart of an aircraft. A modern aeroengine has a complex structure, bad operation environment, fast rotational speed, high maintenance cost and so on. To improve management efficiency and reduce maintenance cost, an airline generally manages all aeroengines with the same model uniformly. In this dissertation, the fleet is defined as all aeroengines with the same model in an airline. From the perspective of an airline, studying aeroengines in a fleet oriented maintenance programming is essential for improving an airline's maintenance management and competitiveness. Based on the analysis of existing problems in studying aeroengines in a fleet oriented maintenance programming at home and abroad, a system of aeroengines in a fleet oriented maintenance programming, which includes maintenance data management, removal deadline prediction, removal scheduling, maintenance workscope decision-making, maintenance cost prediction, maintenance progress monitoring and maintenance effect evaluation is established and its related models and key technologies are studied, according to domestic airline's needs in aeroengine maintenance management.Aeroengine maintenance data management is the basis for aeroengines in a fleet oriented maintenance programming. According to its multi-dimensional, scalable and evolutionary characteristics, aeroengine maintenance data is divided into configuration data, object-related data and class-related data with the help of object-oriented thinking. In the organization of aeroengine maintenance data, usage state based aeroengine maintenance data organization model are established to achieve a unified management of aeroengine maintenance data. In the management of aeroengine configuration data, position part based aeroengine maintenance BOM management model and physical state based instance maintenance BOM management model are established to solve data redundancy, and major part based Airworthiness Directive (AD) / Service Bulletin (SB) status control method is put forward to solve AD / SB status difficult to control. Each aeroengine removal deadline prediction in a fleet is the premise for scheduling aeroengine removal time reasonably. To predict the aeroengine removal deadline more accurately, a prediction method based on aeroengine performance deterioration pattern is put forward considering the differences of each aeroengine's performance deterioration in a fleet. A multi-objective combinatorial optimization model of aeroengines in a fleet oriented removal scheduling is constructed after analyzing influencing factors of aeroengine removal date and spare aeroengine selection, and then a heuristic algorithm based on the progressive structure is put forward to solve the model. The method of the solution set construction and selection for removal scheduling problem is also studied. Finally, the practical data is adopted to verify the effectiveness of this algorithm.When an aeroengine has a shop vist, objectives of shop visit will be always given based on the overall situation of the fleet by the airline. To make a reasonable aeroengine maintenance workscope, an objective-oriented method for aeroengine maintenance workscope decision-making is proposed after analyzing objectives of aeroengine shop visit. Aeroengine maintenance workscope is made through two steps from five aspects of life limited part, component damage, soft time, AD / SB and exhaust gas temperature margin using the method. To solve the problem that module performance restoration value is difficult to distribute in the decision-making process, optimization models based on minimum shop visit cost with certain or uncertain condition are put forward, and dynamic programming and heuristic algorithm are respectively studied to solve optimization models. By comparing their results, the applicable environment of the two algorithms is given. Therefore, aeroengine maintenance workscope decision-making becomes more scientific.Each aeroengine shop visit cost prediction in a fleet is the premise for making annual aeroengine maintenance budget. To predict it accurately, aeroengine shop visit cost is analyzed and divided into the determinacy cost of typeâ… , the determinacy cost of typeâ…¡, the indeterminacy cost of typeâ… and the indeterminacy cost of typeâ…¡. On this basis, aeroengine shop visit cost prediction is studied with having historical cost breakdown data or not. When having historical cost breakdown data, the scrap probability model of important part and the cost probability model of other material and repair are established. When having no historical cost breakdown data, a prediction model named information diffusion support vector machine is proposed. The topology, modeling process and selection of parameters of the model are also studied. The practical data is adopted to verify the effectiveness of models with having historical cost breakdown data or not.To ensure that an aeroengine for shop visit leaves a MRO company on time, aeroengine maintenance work breakdown structure and network planning are presented, and PERT is adopted to monitor aeroengine maintenance progress. To solve existing problems in acquiring the task duration using the traditional PERT, a new multi-point approximation method is presented, and then obtained aeroengine maintenance completion time distribution function using an approximate method. To provide decision-making support for selection of aeroengine MRO company, the aeroengine maintenance effect evaluation system is established, and then data envelopment analysis is adopted to achieve the evaluation of each aeroengine maintenance effect in a fleet. Application cases for aeroengine maintenance progress monitoring and maintenance effect evaluation are presented to verify the effectiveness of methods.Finally, the application of aeroengines in a fleet oriented maintenance programming and its key technologies at a large-scale airline is discussed, by developing an aeroengine health management and maintenance decision-making support system for Air China. Based on analyzing aeroengine maintenance management requirements of Air China, the system architecture, the function model and the information model are introduced. Operation examples and application results of the system are also presented, demonstrating that the proposed models and technologies are correct and effective.