Research For Life On Wing Prediction And Maintenance Decision-Making In Civil Aeroengine Based On CBM Strategy

For airlines, a scientific and reasonable method of civil aeroengine maintenance decision-making should be drew up before making the plan of engine maintenance and removal. Originally,the modeling studies of maintenance in civil aeroengine were commonly based on the statistical analysis of failure life data,and the effects of engine's necessary state informations on the operation of the system were ignored, so the maintenance decisionmaking was largely independent of the system state. Hence,based on the timing issues of engine maintenance decision-making, the research for life on wing prediction and maintenance decision-making in civil aeroengine, according to the currently widely used CBM strategy,was mainly conducted in this paper.The basic theories of maintenance decision-making modeling in civil aeroengine based on CBM strategy were deeply studied in this article,and besides,the model of maintenance decision-making control limit based on proportional hazards-proportional odds(PH-PO)model and the model of state parameters trend prediction based on SVM theory were also established. For engines that were removed due to the performance degradation, the modeling method of engine maintenance decision-making based on CBM strategy was studied, and respectively from the single performance parameter and multiple state parameters. When only one single performance parameter was considered, such as exhaust gas temperature margin(EGTM), some simple trend analysis of performance degradation that used the model of state parameters trend prediction based on SVM theory was firstly conducted,and then the preset monitoring threshold was consulted to determine whether the engine needed a repair.For civil aeroengine which is a very complex system,although the method above is simple,it is relatively rough and there will be some difficulties to reflect the comprehensive effect of multiple influencing factors. Therefore, it is necessary to consider of using a method of comprehensive performance assessment based on multiple state parameters. Under the condition of considering multiple state parameters, the PH-PO model was adopted. Through the research and analysis of historical maintenance data of the engine fleet,multiple stateparameters were selected, and the time on wing of removal due to the performance degradation was regarded as life data samples,and at the same time full consideration about the effects of engine's necessary state informations on the operation of the system was given.As a result,a comparatively accurate mathematic relation was established among the system state parameters,system reliability and the time on wing,and then the model of engine maintenance decision-making was obtained. After that,combined with the trend prediction of engine state parameters that used the model of state parameters trend prediction based on SVM theory,life on wing of engine was predicted.Through the research of this article,the life on wing of civil aeroengine can be predicted dynamically,and in this prediction process,system reliability, system state and fault data were all integrateded. Thus one more flexible method of engine life prediction for airlines and technical support for the plan of engine maintenance and removal were provided. And in the case study of a concrete instance, the result shows the high engineering practical value of the research.