RUL prediction for IMA based on deep regression method

Integrated modular avionics (IMA) is one of the most significant systems which can affect the performance of civil aircraft. The structure of IMA is an open standard platform, on which amount of functionalities are hosted. This structure makes avionics more flexible and more powerful, while it makes avionics more complexity and more critical. For improving the safety and reliability of aircraft, a deep regression method is utilized for predicting the remaining useful life (RUL) of IMA. The deep regression method adopts stacked denoising autoencoders (SDAE) to extract deep features and further improve the quality of the learned features. Then the deep features are fed into support vector machine (SVM) to predict the RUL. An IMA degradation model which accompanied with intermittent faults (IFs) is built based on probability distribution approaches. The simulation is realized by Monte Carlo method. The deep regression method is applied to the RUL prediction of IMA. The results confirm that the deep regression method is more effective and robust compared with the existing methods.