The Research Of Health Management Of Aviation Engine And Engine Fleet Health Assessment Methods

With the rapid development of air transport, aircraft engine health management is particularly important. At the same time, the oil system is an important system in aero engine system and it will have a significant impact to its health status and economic security. Therefore, the oil system health management is important to improve flight safety, improve system performance and reduce maintenance costs. The core problem to resolve for the Health management system is how to realize the system of health assessment and failure prediction. This paper chose the aviation engine oil system as a starting point, proposed the implementation of programs to improve health management systems assessment, forecasting function module, and depend on this to go deep into study system health evaluation and failure prediction of core issues and key technologies.First this paper has study the oil system health assessment method based on SOM neural network. Using SOM neural network to determine the health status of the system under the different parameters and compared it with health indices under fault conditions and divided the health system into different levels providing the basis to judge the health of the oil system in time. Using the fault samples and normal samples to validate on SOM neural network and the results proved the validity of the model. Second, focus on the performance of trend forecasting based on support vector regression model, which was used in aviation engine oil system of trend forecasting. Using genetic algorithms to select the model parameters is effectively preventing the emergence of the phenomenon of over-fitting. Experimental results show that the model can accurately determine the follow-up time of the engine oil system and achieve the engine condition monitoring, fault detection and performance parameters for trend analysis.Finally, by using gray correlation analysis and AHP, this paper proposed a model to assess the health status of aviation engines. The model use the improved AHP to determine the weights of evaluation index and use gray relational analysis evaluated the gray correlation coefficient, the calculated as a weighted gray correlation to determine the health status indicators of the aviation engines. The paper also use SOM neural network and AHP, build health assessment model, get the health status indicators. Based on aircraft engine QAR(Quick access recorder) data, the two model are applied to analyzing the examples of multiple engine health status, the results demonstrate the effectiveness of the method.This paper have study the Engine oil system health assessment and failure prediction method to provide technical support for the aero-engine health management system implementation.