Design controller for synchronization of an array of delayed neural networks using a controllable probabilistic PSO

  • Yang Tang, Zidong Wanga, Jian-an Fanga
  • Published 2010

Abstract

This paper introduces a controllable probabilistic particle swarm optimization (CPPSO) algorithm based on Bernoulli stochastic variables and a competitive penalized method. The CPPSO is proposed to solve optimization problems and is applied to design the memoryless feedback controller, which is used in synchronization of an array of delayed neural networks (DNNs). The learning strategies occur in a random way and are governed by Bernoulli stochastic variables. The expectation of Bernoulli stochastic variables are automatically controlled by search environment. The proposed method not only keeps the diversity of the swarm, but also maintains rapid convergence of the PSO according to a competitive penalized mechanism. In addition, the convergence speed is improved because the inertia weight each particle is automatically computed according to the feedback of fitness value. The efficiency of the proposed CPPSO is demonstrated by comparing it with some well-known PSO algorithms on benchmark test functions with and without rotation. In the end, the proposed CPPSO algorithm is used to design the controller for synchronization of an array of continuous-time delayed neural networks.

11 Figures and Tables

Cite this paper

@inproceedings{Tang2010DesignCF, title={Design controller for synchronization of an array of delayed neural networks using a controllable probabilistic PSO}, author={Yang Tang and Zidong Wanga and Jian-an Fanga}, year={2010} }