An Improved BP Algorithm and its Application

Abstract:

Aiming at BP algorithm convergence slow and be prone to plunge a partial basis, an improved BP algorithm------ACBP Algorithm is proposed in this paper, it has better diversity and global search capacity. The ability of optimization for the algorithm is tested through numerical computation, the experimental demonstrates that the improved BP algorithm has better diversity and global search capacity than genetic algorithm, BP algorithm, ant colony algorithm and simulated anneal algorithm.

• Periodical: Advanced Materials Research (Volumes 718-720)
• Pages: 2026-2029
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.718-720.2026
• Online since: July 2013
• Authors: Ju Ping Peng, Yue Guang Li
• Keywords: ACBP Algorithm, Ant Colony Algorithm (ACA), BP Algorithm, Function Problem

References

[1] Li Minqiang, Xu Boyi. Combination of genetic algorithms and neural networks [J]. systems engineering theory & practice, 1999, 17 ( 4): 37-40.

[2] Zhu Bangtai, Yang Xiaoyu, Zhang Ziqiang. A kind generalization algorithm of BP network [J]. Journal of Luoyang Institute of technology, 1998, 19 (2): 73-76.

[3] Blum C, Dorigo M. The hyper-cube framework for ant colony optimization[C].IEEE Transactions on Systems, Man and Cybernetics.2004.

DOI: 10.1109/tsmcb.2003.821450

[4] He Yubin, Li Xinzhong. Neural network control technology and its application. Science Press, 2000.

[5] Yang Dali, Liu Zemin. The multilayer forward neural network algorithm based on the error analysis and improvement algorithm. Journal of Electronics Vol.23.No.1.117-120,1995.

[6] HAN K-H K-M J-H.Quantum-inspired evolutionary algorithms with a new termination criterion, H, gate, and two-phase scheme [J].IEEE Trans on EvolutionaryComputation,2004,8(2):156-169.

DOI: 10.1109/tevc.2004.823467