中国物理B ›› 2008, Vol. 17 ›› Issue (4): 1196-1201.doi: 10.1088/1674-1056/17/4/008

• • 上一篇    下一篇

Parameters estimation online for Lorenz system by a novel quantum-behaved particle swarm optimization

李卓球1, 高飞2, 童恒庆2   

  1. (1)Department of Engineering and Mechanics, School of Science, Wuhan University of Technology, Wuhan 430070, China; (2)Department of Science, School of Science, Wuhan University of Technology, Wuhan 430070, China
  • 收稿日期:2007-04-16 修回日期:2007-10-08 出版日期:2008-04-20 发布日期:2008-04-01
  • 基金资助:
    Project supported by the National Natural Science Foundation of China (Grant No 10647141).

Parameters estimation online for Lorenz system by a novel quantum-behaved particle swarm optimization

Gao Fei(高飞)a)†, Li Zhuo-Qiu(李卓球)b), and Tong Heng-Qing(童恒庆)a)   

  1. a Department of Science, School of Science, Wuhan University of Technology, Wuhan 430070, China; b Department of Engineering and Mechanics, School of Science, Wuhan University of Technology, Wuhan 430070, China
  • Received:2007-04-16 Revised:2007-10-08 Online:2008-04-20 Published:2008-04-01
  • Supported by:
    Project supported by the National Natural Science Foundation of China (Grant No 10647141).

PDF (PC)

936

赞

摘要： This paper proposes a novel quantum-behaved particle swarm optimization (NQPSO) for the estimation of chaos' unknown parameters by transforming them into nonlinear functions' optimization. By means of the techniques in the following three aspects: contracting the searching space self-adaptively; boundaries restriction strategy; substituting the particles' convex combination for their centre of mass, this paper achieves a quite effective search mechanism with fine equilibrium between exploitation and exploration. Details of applying the proposed method and other methods into Lorenz systems are given, and experiments done show that NQPSO has better adaptability, dependability and robustness. It is a successful approach in unknown parameter estimation online especially in the cases with white noises.

关键词: parameter estimation online, chaos system, quantum particle swarm optimization

Abstract: This paper proposes a novel quantum-behaved particle swarm optimization (NQPSO) for the estimation of chaos' unknown parameters by transforming them into nonlinear functions' optimization. By means of the techniques in the following three aspects: contracting the searching space self-adaptively; boundaries restriction strategy; substituting the particles' convex combination for their centre of mass, this paper achieves a quite effective search mechanism with fine equilibrium between exploitation and exploration. Details of applying the proposed method and other methods into Lorenz systems are given, and experiments done show that NQPSO has better adaptability, dependability and robustness. It is a successful approach in unknown parameter estimation online especially in the cases with white noises.

Key words: parameter estimation online, chaos system, quantum particle swarm optimization

中图分类号:  (Numerical optimization)

  • 02.60.Pn
03.67.Hk (Quantum communication) 03.67.Lx (Quantum computation architectures and implementations) 05.45.-a (Nonlinear dynamics and chaos)