JZUS - Journal of Zhejiang University SCIENCE

Frontiers of Information Technology & Electronic Engineering 2016 Vol.17 No.12 P.1287-1304

Finite-sensor fault-diagnosis simulation study of gas turbine engine using information entropy and deep belief networks

Author(s): De-long Feng, Ming-qing Xiao, Ying-xi Liu, Hai-fang Song, Zhao Yang, Ze-wen Hu
Affiliation(s): Aeronautics and Astronautics Engineering College, Air Force Engineering University, Xi’an 710038, China; more
Corresponding email(s): fengdelong101@foxmail.com
Key Words: Deep belief networks (DBNs), Fault diagnosis, Information entropy, Engine

Share this article to： More <<< Previous Article \|Next Article >>>

Abstract: Precise fault diagnosis is an important part of prognostics and health management. It can avoid accidents, extend the service life of the machine, and also reduce maintenance costs. For gas turbine engine fault diagnosis, we cannot install too many sensors in the engine because the operating environment of the engine is harsh and the sensors will not work in high temperature, at high rotation speed, or under high pressure. Thus, there is not enough sensory data from the working engine to diagnose potential failures using existing approaches. In this paper, we consider the problem of engine fault diagnosis using finite sensory data under complicated circumstances, and propose deep belief networks based on information entropy, IE-DBNs, for engine fault diagnosis. We first introduce several information entropies and propose joint complexity entropy based on single signal entropy. Second, the deep belief networks (DBNs) is analyzed and a logistic regression layer is added to the output of the DBNs. Then, information entropy is used in fault diagnosis and as the input for the DBNs. Comparison between the proposed IE-DBNs method and state-of-the-art machine learning approaches shows that the IE-DBNs method achieves higher accuracy.

基于信息熵和深度置信网络的涡轮发动机在有限传感器下的故障诊断仿真研究

概要：精确故障诊断是预测与健康管理的一个重要部分。它能避免事故的发生，延长设备使用寿命，还能降低设备维修保养费用。本文研究涡轮发动机的故障诊断。由于发动机工作在高温、高压、高转速的严峻环境中，不能安装过多传感器，因此我们无法获得足够多的传感器数据，以至于采用现有算法不能进行精确的潜在故障诊断。本文针对复杂环境下有限传感器数据的发动机故障诊断问题，提出了一种基于信息熵的深度置信网络方法。首先介绍了几种信息熵，并基于单信号熵提出了联合复杂信息熵。其次，分析了深度置信网络的构成，提出了基于信息熵的深度置信网络方法。验证实验表明，与现有的机器学习算法比较，该方法的诊断精度大大提高。

关键词：深度置信网络；信息熵；故障诊断；发动机

Reference

[1]Aguiar, V., Guedes, I., 2015. Shannon entropy, Fisher information and uncertainty relations for log-periodic oscillators. Phys. A, 423:72-79.

[2]Bengio, Y., 2009. Learning Deep Architectures for AI. Available from http://www.iro.umontreal.ca/~bengioy/ papers/ftml.pdf

[3]Bengio, Y., 2012. Practical recommendations for gradient- based training of deep architectures. LNCS, 7700:437- 478.

[4]Bengio, Y., Courville, A., Vincent, P., 2013. Representation learning: a review and new perspectives. IEEE Trans. Patt. Anal. Mach. Intell., 35(8):1798-1828.

[5]Bottou, L., 2012. Stochastic gradient descent tricks. LNCS, 7700:421-436.

[6]Chen, Y.S., Zhao, X., Jia, X.P., 2015. Spectral-spatial classification of hyperspectral data based on deep belief network. IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens., 8(6):2381-2392.

[7]Cui, H.X., Zhang, L.B., Kang, R.Y., et al., 2009. Research on fault diagnosis for reciprocating compressor valve using information entropy and SVM method. J. Loss Prevent. Process Ind., 22(6):864-867.

[8]Dai, J.H., Tian, H.W., 2013. Entropy measures and granularity measures for set-valued information systems. Inform. Sci., 240:72-82.

[9]Ferrer, A., 2007. Multivariate statistical process control based on principal component analysis (MSPC-PCA): some reflections and a case study in an autobody assembly process. Qual. Eng., 19(4):311-325.

[10]Geng, J.B., Huang, S.H., Jin, J.S., et al., 2006. A method of rotating machinery fault diagnosis based on the close degree of information entropy. Int. J. Plant Eng. Manag., 11(3):137-144.

[11]Hinton, G.E., 2010. A Practical Guide to Training Restricted Boltzmann Machines. Available from https://www.cs. toronto.edu/~hinton/absps/guideTR.pdf

[12]Hinton, G.E., Osindero, S., Teh, Y.W., 2006. A fast learning algorithm for deep belief nets. Neur. Comput., 18(7): 1527-1554.

[13]Hinton, G.E., Deng, L., Yu, D., et al., 2012. Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups. IEEE Signal Process. Mag., 29(6):82-97.

[14]Jin, C.X., Li, F.C., Li, Y., 2014. A generalized fuzzy ID3 algorithm using generalized information entropy. Knowl.- Based Syst., 64:13-21.

[15]Koverda, V.P., Skokov, V.N., 2012. Maximum entropy in a nonlinear system with a 1/f power spectrum. Phys. A, 391(1-2):21-28.

[16]Larochelle, H., Bengio, Y., Louradour, J., et al., 2009. Exploring strategies for training deep neural networks. J. Mach. Learn. Res., 10(10):1-40.

[17]Li, F.C., Zhang, Z., Jin, C.X., 2016. Feature selection with partition differentiation entropy for large-scale data sets. Inform. Sci., 329:690-700.

[18]Li, J., 2015. Recognition of the optical image based on the wavelet space feature spectrum entropy. Optik-Int. J. Light Electron Opt., 126(23):3931-3935.

[19]Liu, Z.G., Hu, Q.L., Cui, Y., et al., 2014. A new detection approach of transient disturbances combining wavelet packet and Tsallis entropy. Neurocomputing, 142:393- 407.

[20]Martens, J., Sutskever, I., 2012. Training deep and recurrent networks with Hessian-free optimization. LCNS, 7700: 479-535.

[21]Memisevic, R., Hinton, G.E., 2010. Learning to represent spatial transformations with factored higher-order Boltzmann machine. Neur. Comput., 22(6):1473-1492.

[22]Mohamed, A.R., Dahl, G.E., Hinton, G.E., 2012. Acoustic modeling using deep belief networks. IEEE Trans. Audio Speech Lang. Process., 20(1):14-22.

[23]Nichols, J.M., Seaver, M., Trickey, S.T., 2006. A method for detecting damage-induced nonlinearities in structures using information theory. J. Sound Vibr., 297(1-2):1-16.

[24]Niu, J., Bu, X.Z., Li, Z., et al., 2014. An improved bilinear deep belief network algorithm for image classification. 10th Int. Conf. on Computational Intelligence and Security, p.189-192.

[25]Nourani, V., Alami, M.T., Vousoughi, F.D., 2015. Wavelet- entropy data pre-processing approach for ANN-based groundwater level modeling. J. Hydrol., 524:255-269.

[26]Ong, B.T., Sugiura, K., Zettsu, K., 2014. Dynamic pre-training of deep recurrent neural networks for predicting environmental monitoring data. IEEE Int. Conf. on Big Data, p.760-765.

[27]Pan, Y.B., Yang, B.L., Zhou, X.W., 2015. Feedstock molecular reconstruction for secondary reactions of fluid catalytic cracking gasoline by maximum information entropy method. Chem. Eng. J., 281:945-952.

[28]Rastegin, A.E., 2015. On generalized entropies and information-theoretic Bell inequalities under decoherence. Ann. Phys., 355:241-257.

[29]Rodríguez, P.H., Alonso, J.B., Ferrer, M.A., et al., 2013. Application of the Teager-Kaiser energy operator in bearing fault diagnosis. ISA Trans., 52(2):278-284.

[30]Saimurugan, M., Ramachandran, K.I., Sugumaran, V., et al., 2011. Multi component fault diagnosis of rotational mechanical system based on decision tree and support vector machine. Expert Syst. Appl., 38(4):3819-3826.

[31]Sainath, T.N., Kingsbury, B., Soltau, H., et al., 2013. Optimization techniques to improve training speed of deep neural networks for large speech tasks. IEEE Trans. Audio Speech Lang. Process., 21(11):2267-2276.

[32]Sainath, T.N., Kingsbury, B., Saon, G., et al., 2015. Deep convolutional neural networks for large-scale speech tasks. Neur. Networks, 64:39-48.

[33]Sekerka, R.F., 2015. Entropy and information theory. In: Thermal Physics: Thermodynamics and Statistical Mechanics for Scientists and Engineers. Elsevier, p.247-256.

[34]Sermanet, P., Chintala, S., LeCun, Y., 2012. Convolutional neural networks applied to house numbers digit classification. 21st Int. Conf. on Pattern Recognition, p.3288- 3291.

[35]Song, X.D., Sun, G.H., Dong, S.H., 2015. Shannon information entropy for an infinite circular well. Phys. Lett. A, 379(22-23):1402-1408.

[36]Su, H.T., You, G.J.Y., 2014. Developing an entropy-based model of spatial information estimation and its application in the design of precipitation gauge networks. J. Hydrol., 519(D):3316-3327.

[37]Susan, S., Hanmandlu, M., 2013. A non-extensive entropy feature and its application to texture classification. Neurocomputing, 120:214-225.

[38]Sutskever, I., Hinton, G.E., Taylor, G.W., 2008. The recurrent temporal restricted Boltzmann machine. Proc. 22nd Annual Conf. on Neural Information Processing Systems, p.1601-1608.

[39]Tamilselvan, P., Wang, P.F., 2013. Failure diagnosis using deep belief learning based health state classification. Reliab. Eng. Syst. Safety, 115:124-135.

[40]Tamilselvan, P., Wang, P.F., Youn, B.D., 2011. Multi-sensor health diagnosis using deep belief network based state classification. ASME Int. Design Engineering Technical Conf. & Computers and Information in Engineering Conf., p.749-758.

[41]Tran, V.T., AlThobiani, F., Ball, A., 2014. An approach to fault diagnosis of reciprocating compressor valves using Teager–Kaiser energy operator and deep belief networks. Expert Syst. Appl., 41(9):4113-4122.

[42]Xie, Y., Zhang, T., 2005. A fault diagnosis approach using SVM with data dimension reduction by PCA and LDA method. Chinese Automation Congress, p.869-874.

[43]Zhang, W.L., Li, R.J., Deng, H.T., et al., 2015. Deep convolutional neural networks for multi-modality isointense infant brain image segmentation. NeuroImage, 108:214- 224.

[44]Zhao, X.Z., Ye, B.Y., 2016. Singular value decomposition packet and its application to extraction of weak fault feature. Mech. Syst. Signal Process., 70-71:73-86.

[45]Zhou, S.S., Chen, Q.C., Wang, X.L., 2014. Deep adaptive networks for visual data classification. J. Multim., 9(10): 1142-1151.

Open peer comments: Debate/Discuss/Question/Opinion

<1>

Similar articles

- Go to

基于信息熵和深度置信网络的涡轮发动机在有限传感器下的故障诊断仿真研究

Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article

Reference