论文信息 - Aero-engine Sensor Fault Diagnosis Based on Convolutional Neural Network

Aero-engine Sensor Fault Diagnosis Based on Convolutional Neural Network

In this paper, a deep model based on Convolutional Neural Network is proposed to solve the problem of sensor fault diagnosis for aero-engine. The model uses multiple convolutional layers to extract the features of the fault signal of the aero-engine sensor and uses the Softmax classifier to diagnose the fault. Through the comparison and analysis of the experiment, the accuracy of this method in fault diagnosis of aero-engine sensors can reach 100%. On the basis of this, the fault diagnosis method base on PCA-CNN is further proposed. Firstly, PCA is used to reduce the dimension of the original data, and then construct a Convolutional Neural Network for fault diagnosis. This method greatly improves the efficiency of the algorithm compared with CNN.

Weidong Qu | Jian Li

[1] Sergey Ioffe,et al. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[2] Adrian G. Bors,et al. Principal Component Analysis of spectral coefficients for mesh watermarking , 2008, 2008 15th IEEE International Conference on Image Processing.

[3] Rob Fergus,et al. Visualizing and Understanding Convolutional Neural Networks , 2013 .

[4] Yee Whye Teh,et al. A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[5] Yoshua Bengio,et al. Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[6] Yan Bing,et al. Aero-engine sensor fault diagnosis based on stacked denoising autoencoders , 2016, 2016 35th Chinese Control Conference (CCC).

[7] Wang Le. Research on ELM-based aircraft engine sensor fault diagnosis method , 2015 .