Generator rotating rectifier fault detection method based on stacked auto-encoder

This paper presents a method of fault features extraction based on the stacked auto-encoder (SAE), and this method can be used to rotating rectifier diodes open-circuit fault detection of brushless ac synchronous generator. First, exciter generator field current is collected, and then processed by Fast Fourier Transform (FFT) to obtain the frequency components. Secondly, these components are deeply learned and evaluated by the stacked auto-encoder to extract features automatically. Finally, fault detection can be carried out with Euclidean distance calculation. The algorithm is validated based on a brushless ac synchronous generator test rig, and the experimental results show perfect fault detection performance under different load conditions.

[1]  Abdelkader Chaari,et al.  Brushless Three-Phase Synchronous Generator Under Rotating Diode Failure Conditions , 2014, IEEE Transactions on Energy Conversion.

[2]  Yee Whye Teh,et al.  Unsupervised Discovery of Nonlinear Structure Using Contrastive Backpropagation , 2006, Cogn. Sci..

[3]  Abdelkader Chaari,et al.  Detection of brushless exciter rotating diodes failures by spectral analysis of main output voltage , 2013, 2013 International Conference on Electrical Engineering and Software Applications.

[4]  D.J. Morrow,et al.  Noninvasive detection of brushless exciter rotating diode failure , 2004, IEEE Transactions on Energy Conversion.

[5]  Antonio Griffo,et al.  Sensorless starting of a wound-field synchronous starter/generator for aerospace applications , 2012, IEEE Transactions on Industrial Electronics.

[6]  Nathan Kumbar,et al.  Advanced Electrical Signature Analysis of Aircraft Electrical Generators , 2009 .

[7]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[8]  Gautam Biswas,et al.  Aircraft Power Generators: Hybrid Modeling and Simulation for Fault Detection , 2012, IEEE Transactions on Aerospace and Electronic Systems.

[9]  Stephen Lin,et al.  Automatic Feature Learning to Grade Nuclear Cataracts Based on Deep Learning , 2014, IEEE Transactions on Biomedical Engineering.

[10]  T.D. Batzel,et al.  Prognostic Health Management of Aircraft Power Generators , 2009, IEEE Transactions on Aerospace and Electronic Systems.

[11]  Xiang Bai,et al.  Vehicle Color Recognition With Spatial Pyramid Deep Learning , 2015, IEEE Transactions on Intelligent Transportation Systems.

[12]  Nishchal K. Verma,et al.  Generating feature sets for fault diagnosis using denoising stacked auto-encoder , 2016, 2016 IEEE International Conference on Prognostics and Health Management (ICPHM).

[13]  Sanjib Kumar Panda,et al.  Feasibility analysis of auxiliary winding for condition monitoring of wound field brushless synchronous generators , 2015, IECON 2015 - 41st Annual Conference of the IEEE Industrial Electronics Society.