An integrated approach to planetary gearbox fault diagnosis using deep belief networks

Aiming at improving the accuracy of planetary gearbox fault diagnosis, an integrated scheme based on dimensionality reduction method and deep belief networks (DBNs) is presented in this paper. Firstly, the acquired vibration signals are decomposed into mono-component called intrinsic mode functions (IMFs) through ensemble empirical mode decomposition (EEMD), and then Teager–Kaiser energy operator (TKEO) is used to track the instantaneous amplitude (IA) and instantaneous frequency (IF) of a mono-component amplitude modulation (AM) and frequency modulation (FM) signal. Secondly, a high dimensional feature set is constructed through extracting statistical features from six different signal groups. Then, an integrated dimensionality reduction method combining feature selection and feature extraction techniques is proposed to yield a more sensitive and lower dimensional feature set, which not only reduces the computation burden for fault diagnosis but also improves the separability of the samples by integrating the label information. Further, the low dimensional feature set is fed into DBNs classifier to identify the fault types using the optimal parameters selected by particle swarm optimization algorithm (PSO). Finally, two independent cases study of planetary gearbox fault diagnosis are carried out on test rig, and the results show that the proposed method provides higher accuracy in comparison with the existing methods.

[1]  Robert X. Gao,et al.  Rotary Machine Health Diagnosis Based on Empirical Mode Decomposition , 2008 .

[2]  Abdolreza Ohadi,et al.  Comparison of FDA-based and PCA-based features in fault diagnosis of automobile gearboxes , 2013, Neurocomputing.

[3]  Yaguo Lei,et al.  A review on empirical mode decomposition in fault diagnosis of rotating machinery , 2013 .

[4]  Geoffrey E. Hinton.  A Practical Guide to Training Restricted Boltzmann Machines , 2012, Neural Networks: Tricks of the Trade.

[5]  Ming J. Zuo,et al.  Joint amplitude and frequency demodulation analysis based on local mean decomposition for fault diagnosis of planetary gearboxes , 2013 .

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

[7]  Kaixiang Peng,et al.  Adaptive total PLS based quality-relevant process monitoring with application to the Tennessee Eastman process , 2015, Neurocomputing.

[8]  Yaguo Lei,et al.  Fault detection of planetary gearboxes using new diagnostic parameters , 2012 .

[9]  Zhipeng Feng,et al.  Iterative generalized synchrosqueezing transform for fault diagnosis of wind turbine planetary gearbox under nonstationary conditions , 2015 .

[10]  Dong Han,et al.  Planetary gearbox fault diagnosis using an adaptive stochastic resonance method , 2013 .

[11]  Ming J. Zuo,et al.  Fault level diagnosis for planetary gearboxes using hybrid kernel feature selection and kernel Fisher discriminant analysis , 2013 .

[12]  Ming J. Zuo,et al.  Classification of gear damage levels in planetary gearboxes , 2011, 2011 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications (CIMSA) Proceedings.

[13]  K. I. Ramachandran,et al.  Feature selection using Decision Tree and classification through Proximal Support Vector Machine for fault diagnostics of roller bearing , 2007 .

[14]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[15]  Balbir S. Dhillon,et al.  Early fault diagnosis of rotating machinery based on wavelet packets—Empirical mode decomposition feature extraction and neural network , 2012 .

[16]  Peter J. Bickel,et al.  Maximum Likelihood Estimation of Intrinsic Dimension , 2004, NIPS.

[17]  Ming J. Zuo,et al.  Vibration signal models for fault diagnosis of planetary gearboxes , 2012 .

[18]  Fuhui Long,et al.  Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy , 2003, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[19]  Z. H. Che,et al.  PSO-based back-propagation artificial neural network for product and mold cost estimation of plastic injection molding , 2010, Comput. Ind. Eng..

[20]  Petros Maragos,et al.  On amplitude and frequency demodulation using energy operators , 1993, IEEE Trans. Signal Process..

[21]  Ahmad Ghasemloonia,et al.  Application and comparison of an ANN-based feature selection method and the genetic algorithm in gearbox fault diagnosis , 2011, Expert Syst. Appl..

[22]  Shuo-Yan Chou,et al.  A simulated-annealing-based approach for simultaneous parameter optimization and feature selection of back-propagation networks , 2008, Expert Syst. Appl..

[23]  Yaguo Lei,et al.  Phenomenological models of vibration signals for condition monitoring and fault diagnosis of epicyclic gearboxes , 2016 .

[24]  Pingfeng Wang,et al.  Failure diagnosis using deep belief learning based health state classification , 2013, Reliab. Eng. Syst. Saf..

[25]  Haidong Shao,et al.  Rolling bearing fault diagnosis using an optimization deep belief network , 2015 .

[26]  Radoslaw Zimroz,et al.  Vibration condition monitoring of planetary gearbox under varying external load , 2009 .

[27]  Yaguo Lei,et al.  New clustering algorithm-based fault diagnosis using compensation distance evaluation technique , 2008 .

[28]  I. Soltani Bozchalooi,et al.  Teager energy operator for multi-modulation extraction and its application for gearbox fault detection , 2010 .

[29]  Byeng D. Youn,et al.  Multi-Sensor Health Diagnosis Using Deep Belief Network Based State Classification , 2011 .

[30]  Pascal Vincent,et al.  Representation Learning: A Review and New Perspectives , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[31]  Yaguo Lei,et al.  Condition monitoring and fault diagnosis of planetary gearboxes: A review , 2014 .

[32]  Michael N. Vrahatis,et al.  Recent approaches to global optimization problems through Particle Swarm Optimization , 2002, Natural Computing.

[33]  Hao Tian,et al.  A new feature extraction and selection scheme for hybrid fault diagnosis of gearbox , 2011, Expert Syst. Appl..

[34]  Xin Zhou,et al.  Deep neural networks: A promising tool for fault characteristic mining and intelligent diagnosis of rotating machinery with massive data , 2016 .

[35]  Andrew D. Ball,et al.  An approach to fault diagnosis of reciprocating compressor valves using Teager-Kaiser energy operator and deep belief networks , 2014, Expert Syst. Appl..

[36]  ZhiQiang Chen,et al.  Multi-layer neural network with deep belief network for gearbox fault diagnosis , 2015 .

[37]  Darryll J. Pines,et al.  A review of vibration-based techniques for helicopter transmission diagnostics , 2005 .

[38]  Zhipeng Feng,et al.  Fault diagnosis for wind turbine planetary gearboxes via demodulation analysis based on ensemble empirical mode decomposition and energy separation , 2012 .

[39]  Robert B. Randall,et al.  Application of spectral kurtosis for detection of a tooth crack in the planetary gear of a wind turbine , 2009 .

[40]  Cong Wang,et al.  Construction of hierarchical diagnosis network based on deep learning and its application in the fault pattern recognition of rolling element bearings , 2016 .

[41]  Jonathan A. Keller,et al.  Detection of a fatigue crack in a UH-60A planet gear carrier using vibration analysis , 2006 .

[42]  Muhammad Abid,et al.  Exponential discriminant analysis for fault diagnosis , 2016, Neurocomputing.