A gas path fault diagnostic model for gas turbine based on deep belief network with prior information

Gas turbine is widely used in national defense and energy industry. It is significant to carry out gas path fault diagnosis of gas turbine, due to its great impacts on operation and maintenance. With the development of sensor technology and information technology, it is the time of multi-source data generation automatically, so as to provide a good foundation for the development of gas path fault diagnosis. However, the relationship between fault effects and fault modes has not been demonstrated clearly. Thus the theoretical basis for mechanism model is lacked, and data-driven approach for gas path diagnosis is increasingly attractive. In this paper, a novel data-driven model for gas path fault diagnosis based on deep belief network with prior information has been proposed. The effect of network structure on diagnostic accuracy has been studied, and the comparison of this approach between other data-driven approaches has been conducted. The comparing result confirms that this model has an obvious advantage over conventional data-driven models after structure optimization and can be employed to gas path fault diagnosis of gas turbine.

[1]  Huisheng Zhang,et al.  A New Gas Path Fault Diagnostic Method of Gas Turbine Based on Support Vector Machine , 2014 .

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

[3]  Donald L. Simon,et al.  An Integrated Approach for Aircraft Engine Performance Estimation and Fault Diagnostics , 2013 .

[4]  Geoffrey E. Hinton.  Deep belief networks , 2009, Scholarpedia.

[5]  Michele Pinelli,et al.  Artificial Intelligence for the Diagnostics of Gas Turbines: Part II — Neuro-Fuzzy Approach , 2005 .

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

[7]  Zhang Hong-peng.  Development and Electric Power Generation Technology of the Combustion Turbine , 2008 .

[8]  Yoshua. Bengio,et al.  Learning Deep Architectures for AI , 2007, Found. Trends Mach. Learn..

[9]  Dengji Zhou,et al.  SA-PSO Hybrid Algorithm for Gas Path Diagnostics of Gas Turbine , 2016 .

[10]  Yoshua Bengio,et al.  Exploring Strategies for Training Deep Neural Networks , 2009, J. Mach. Learn. Res..

[11]  Mohammadreza Tahan,et al.  Performance-based health monitoring, diagnostics and prognostics for condition-based maintenance of gas turbines: A review , 2017 .

[12]  Huisheng Zhang,et al.  A novel grey prognostic model based on Markov process and grey incidence analysis for energy conversion equipment degradation , 2016 .

[13]  Bernardo Fortunato,et al.  Feed Forward Neural Network-Based Diagnostic Tool for Gas Turbine Power Plant , 2002 .

[14]  Wang Ning,et al.  Gas turbine fault diagnosis based on ART2 Neural Network , 2008, 2008 7th World Congress on Intelligent Control and Automation.

[15]  Christian Igel,et al.  Training restricted Boltzmann machines: An introduction , 2014, Pattern Recognit..

[16]  Feng Jia,et al.  An Intelligent Fault Diagnosis Method Using Unsupervised Feature Learning Towards Mechanical Big Data , 2016, IEEE Transactions on Industrial Electronics.

[17]  Shi Xiao-cheng.  Gas turbine fault diagnosis based on neural network , 2006 .

[18]  Geoffrey E. Hinton.  Training Products of Experts by Minimizing Contrastive Divergence , 2002, Neural Computation.

[19]  Feng Lu,et al.  Gas Path Health Monitoring for a Turbofan Engine Based on a Nonlinear Filtering Approach , 2013 .

[20]  Jie Liu,et al.  Multi-mode diagnosis of a gas turbine engine using an adaptive neuro-fuzzy system , 2017 .

[21]  Khashayar Khorasani,et al.  A component map tuning method for performance prediction and diagnostics of gas turbine compressors , 2014 .