Unknown Parameter Estimation For Chaotic System Based On A Class Of Improved Particle Swarm Optimization

It’s vitally important to estimate the unknown parameter in chaos control and syn-chronization. The application of particle swarm optimization to estimate the unknown parameter of chaotic system through constructing a suitable fitness function, which has become a great research topic. Particle swarm optimization is characterized as a simple concept, which is both easy to implement and compute efficiently. Due to this rea-son, it has attracted great attention and wide application in various fields. While it has shown some disadvantages in specific problems. It sometimes is easy to be trapped in local optimum, and the convergence rate decreased considerably in the later period of evolution; when reaching a near optimal solution, the algorithm stops optimizing, and thus the achieved accuracy of algorithm is limited. In this paper, four improved particle swarm optimizations have been provided, the performance of the proposed algorithms are characterized through estimating the parameter of chaotic system:1.. Through improving the inertia weight> synchronous learning rates and intro-duncing a compressibility factor, three improved adaptive particle swarm optimizations are given:adaptive particle swarm optimization with asynchronous changing learning rates、adaptive particle swarm optimization with compressibility factor、adaptive par-ticle swarm optimization with compressibility factor and asynchronous changing learn-ing rates. In addition, particle swarm optimization and the three improved algorithms are applied to estimating the unknown parameters of Lu system and Lorenz system.2、By introducing drift operator and improving the synchronous learning rates, particle swarm optimization with drift operator and asynchronous changing learning rates is proposed. In addition, particle swarm optimization and the improved algorithm are applied to estimating the unknown parameters of Chen system.As compared with the particle swarm optimization and the modified particle swarm optimization, experimental results show that the improved methods have a higher cal-culation accuracy, and a faster convergence speed.