Aircraft Fault Diagnosis Based on Deep Belief Network

It is a great challenge to accurately and automatically diagnose different faults of aircraft using traditional method. In this paper, a new method based on deep belief network is proposed for aircraft key parts fault diagnosis. Firstly, a deep belief network is constructed with a series of pre-trained restricted Boltzmann machines for feature learning. Secondly, the highest level features learned from the DBN are fed into a softmax classifier for fault diagnosis. Finally, back-propagation learning algorithm is adopted to fine-tune the deep model parameters to further improve the diagnosis accuracy. The proposed method is applied to analyze the experimental rolling bearing signals. The results show that the proposed method is more effective and robust than other traditional methods.