| Summary: | 碩士 === 育達科技大學 === 資訊管理所 === 103 === Today, the demand for various types of aviation aircraft radar and other electronic components for enhancing fighter combat missions has increased. Determining the probability of relative failure of the task or instrument landing system significantly enhances flight safety. The cooling function of the environmental control system for providing cold air to electronic equipment can be considered critical in operating aircraft avionic systems.
In this study, the cooling component of the environmental control system was used in one of the ROCAF fighters as an example. Firstly, the Delphi method was used to collect key factors from experts to determine the most critical component. From the results, it was determined that the most critical component was the “cooling turbine.” Secondly, a five-level Likert scale was used to request experts to evaluate importance and give a score on the key factors that influence the cooling turbine’s life span. Accordingly, four key factors including TSC (time since check), engine high stage bleed air, turbine compressed discharge air ratio, and turbine discharge temperature were selected. According to historical maintenance reports from 2010 to 2014, data of the four key factors were used as input in a BPN (back-propagation network) to train the relationship between input and output to build a prediction mode using the Alyuda Neuro-Intelligence software. The most suitable mode for predicting the next fault moment of the cooling turbine was found to be the mode having one hidden layer with six neurons, 0.1 learning rate, and 10,000 iterations.
The results indicated that the correlation and the R-squared in this model of 0.984129 and 0.963150, respectively, could be obtained after BPN training. Certainly, BPN served as an effective method of predicting the aircraft component’s life.
In this study, the BPN and the time required to construct a critical component was used to predict failure of the environmental control system for developing an optimum predictive model, thereby enabling the component to develop a failure reaction strategy. Hopefully, this strategy will be able to ensure maximum output and minimum input in order to maintain proper aircraft conditions for perfecting military operations and maximizing combat power.
Keywords: Environmental Control System, BPN, Engine High Stage Bleed Air, Turbine Compressed Discharge Temperature.
|
|---|
