Embedded Differential Evolution Algorithm for Recurrent Fuzzy Neural Network Controller Optimization

Abstract:

A chaos concise differential evolution algorithm (CcDE) is proposed for the embedded controller with limited memory, which introduces chaotic local search based on basic differential evolution algorithm to increase exploring and prevent premature convergence. Using virtual population and Gaussian sampling, the CcDE becomes simple and reduces the memory requirements at run time. Experimental simulation on optimizing parameters of the recurrent fuzzy neural network shows that the proposed CcDE can obtain better performance than other concise algorithm.

• Periodical: Applied Mechanics and Materials (Volumes 321-324)
• Pages: 2141-2145
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.321-324.2141
• Online since: June 2013
• Authors: Xiao Sheng Wang, Ying Li, Yan Hui Guo
• Keywords: Chaos Concise Differential Evolution, Embedded Controller, Online Optimization, Recurrent Fuzzy Neural Network, Virtual Population

References

[1] C.F. Hsu and K.H. Cheng: Neurocomputing, Vol. 71(16-18) (2008), pp.3447-3459

[2] J. Ilonen, J.K. Kamarainen, J. Lampinen: Neural Processing Letters, Vol.17(1)(2003),pp.93-105

DOI: 10.1023/a:1022995128597

[3] Y. Zhang, D.W. Gong, Y.Q. Ren, J.H. Zhang: ACTA ELECTRONICA SINICA. （in Chinese）Vol.39(6)(2011), pp.1436-1440

[4] E. Mininno, F. Cupertino, D. Naso: IEEE Trans on Evolutionary Computation. Vol.12(2)(2008), pp.203-219

[5] F. Neri, E. Mininno: IEEE Computational Intelligence Magazine. Vol.5(2)(2010), pp.54-65

[6] R. Caponetto, L. Fortuna, S. Fazzino, M.G. Xibilia: IEEE Transactions on Evolutionary Computation. Vol.7(3)(2003), pp.289-304

DOI: 10.1109/tevc.2003.810069