Closed Loop Dynamic Fuzzy Neural Network for Real-Time Lifetime Forecasting

Abstract:

Real-time lifetime forecasting has extensive application in the fields of machine system manufacturing and integration, which is a good way to promote the dependability and operation stability. In this paper, a closed loop adaptive forecasting model with feedback channel of state monitoring information is built up for the real-time lifetime forecasting. The difference of working state between prediction and monitoring information is used to evaluate the prediction performance. The dynamic fuzzy neural network introduced into the prediction model, in which the fuzzy rule, membrane function and structure parameters can be adjusted according to the evaluate results. A service lifetime testing experiment of gear case is utilized to validate the prediction model. The proposed model achieved reasonable precision with an error of less than 1 hour between the failure time of experimental results and the forecasting remaining lifetime. The adaptive prediction method can deal with the real-time lifetime forecasting for multiple factors and nonlinear system without specific parameters structure.

• Periodical: Advanced Materials Research (Volumes 834-836)
• Pages: 1074-1080
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.834-836.1074
• Online since: October 2013
• Authors: Wen Wang Li, Gao Feng Zheng*, Jian Yi Zheng
• Keywords: Closed Loop, Dynamic Fuzzy, Lifetime Forecasting, Neural Network (NN), Real-Time Prediction

References

[1] S. Momon, M. Moevus, N. Godin, M. R'Mili, P. Reynaud, G. Fantozzi, and G. Fayolle, Composites Part A: Applied Science and Manufacturing, vol. 41, no. 7, pp.913-918, (2010).

DOI: 10.1016/j.compositesa.2010.03.008

[2] U. Reuter and B. Moeller, Computer-Aided Civil and Infrastructure Engineering, vol. 25, no. 5, pp.363-374, (2010).

[3] C. A. M. de Meijer, I. M. Majer, M. A. Koopmanschap, and P. H. van Baal, Medical Care, vol. 50, no. 8, pp.722-729, (2012).

DOI: 10.1097/mlr.0b013e31824ebddc

[4] E. Barrios and J. R. Lansangan, Philippine Management Review, vol. 19, no. 1, pp.23-34, (2012).

[5] M. Noori and M. Ardakani, Lifetime Analysis of Random Event-Driven Clustered Wireless Sensor Networks, Mobile Computing, IEEE Transactions on, vol. 10, no. 10, pp.1448-1458, (2011).

DOI: 10.1109/tmc.2010.254

[6] S. Eryılmaz, Discrete Applied Mathematics, vol. 159, no. 15, pp.1646-1649, (2011).

[7] B. Chen, X. Liu, K. Liu, P. Shi, and C. Lin, Information Sciences, vol. 180, no. 5, pp.776-792, (2010).

[8] X. Deng and X. Wang, Fuzzy Sets and Systems, vol. 160, no. 7, pp.972-987, (2009).

[9] S. Freitag, M. Beer, W. Graf, and M. Kaliske, Computers & Structures, vol. 87, no. 19–20, pp.1187-1194, (2009).

DOI: 10.1016/j.compstruc.2008.12.007

[10] Z. Cai, H. Fang, R. Xiao, R. Xin. A closed-loop supply chain planning based on adaptive genetic algorithm, Industrial Engineering and Engineering Management, 2009 IEEE International Conference on, 1474-1478, (2009).

DOI: 10.1109/ieem.2009.5373071

[11] B. Q. Huang, H. L. Gao, M. H. Xu, X. X. Wu, M. Zhao, L. Guo. Life prediction of CNC linear rolling guide based on DFNN performance degradation model, 2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery (FSKD), 1310-1314, (2009).

DOI: 10.1109/fskd.2010.5569106

[12] A. Lemos, R. Ballini, W. Caminhas, and F. Gomide, Soft Computing: State of the Art Theory and Novel Applications, vol. 291, no. 3, pp.255-268, (2013).

DOI: 10.1007/978-3-642-34922-5_18

[13] J. Yan, C. Guo and X. Wang, A dynamic multi-scale Markov model based methodology for remaining life prediction, Mechanical Systems and Signal Processing, vol. 25, no. 4, pp.1364-1376, (2011).

DOI: 10.1016/j.ymssp.2010.10.018

[14] P. Mastorocostas and C. Hilas, Advances in Neural Networks, vol. 6675, no. 8, pp.529-537, (2011).

[15] R. Kuo, S. Hong and Y. Huang, Applied Mathematical Modelling, vol. 34, no. 12, pp.3976-3990, (2010).