基于改进粒子群优化算法的混沌系统参数估计

Parameter estimation for chaotic system based on improved particle swarm optimization algorithm

提出了一种改进的粒子群优化算法用于解决混沌系统的参数估计问题,从粒子种群的初始化、惯性权重调整策略、差分变异进化、粒子位置与飞行速度的越界处理、局部变尺度深度搜索5个方面对标准粒子群算法进行综合改进,合理有效平衡了算法的全局探索能力与局部开发能力.基准函数测试表明了该算法的全局搜索能力、可靠性及搜索速度都有很大改善,有效克服了标准粒子群算法的早熟收敛现象.以Lorenz混沌系统为例进行仿真实验,结果验证了所提方法的有效性.

Parameter estimation for chaotic systems is a multi-dimensional variable optimization problem,which is one of the key issues in chaotic control and synchronization.An improved particle swarm optimization（IPSO） algorithm was proposed to solve the above mentioned problem.More specifically,the algorithm was comprehensively improved from the aspects of particle population initialization,inertia weight adjustment strategy,differential mutation operation,the boundary violation treatment of particle position and flight velocity as well as local variable-depth search.Therefore,the global exploration and the local exploitation of the algorithm are reasonably and effectively balanced.The experimental results of benchmark function tests show that the global optimization performance,the reliability performance and the search speed are greatly improved.Moreover,the premature convergence of the standard particle swarm algorithm could be effectively avoided.Taking Lorenz chaotic system as an example,the validity of IPSO is also demonstrated by the simulation results.