Computer Engineering and Applications ›› 2008, Vol. 44 ›› Issue (14): 76-79.

• 理论研究 • Previous Articles     Next Articles

Training neural networks with PSO for identification of chaotic systems

WU Zheng-long,WANG Qi,LIU Kai-gang   

  1. Artillery Academy of PLA,Hefei 230031,China
  • Received:2007-09-03 Revised:2007-12-10 Online:2008-05-11 Published:2008-05-11
  • Contact: WU Zheng-long


吴正龙,王 奇,刘开刚   

  1. 解放军炮兵学院 一系,合肥 230031
  • 通讯作者: 吴正龙

Abstract: BP is the most commonly used artificial neural networks,but it suffers from extensive computations,relatively slow convergence speed and other possible weaknesses for complex problems.Genetic Algorithm(GA)has been successfully used to train neural networks,but often with the result of exponential computational complexities and hard implementation.Hence Particle Swarm Optimization(PSO)is used to train BP in the paper.The efficiency of BP trained with PSO is compared with those of BP and BP trained with GA based on the identification of chaotic system.Comparison based on the searching precision and convergence speed of each method shows that BP trained with PSO is dominant and effective to identify chaotic system.

Key words: neural networks, chaotic system, particle swarm optimization, system identification

摘要: 提出利用粒子群优化算法训练神经网络的算法,进行混沌系统辨识,并与神经网络、遗传神经网络对同一混沌系统辨识的结果进行比较。实验表明,利用粒子群优化算法训练神经网络进行混沌系统辨识,在不明显增加执行时间的基础上,寻求最优解的质量有显著提高,并且原理简单,容易实现,可有效用于混沌系统的辨识。

关键词: 神经网络, 混沌系统, 粒子群优化算法, 系统辨识