基于动态参数优化的混沌粒子群算法

Dynamic parameter adjustment based chaos PSO algorithm

针对传统粒子群优化算法易陷入局部极值点的问题,将混沌运动的遍历性,随机性以及初值敏感性等特点融入粒子群优化过程中,并通过模拟退火的方法对参数实现局部优化,使得粒子群优化算法的参数随着优化算法的进行不断改变,以适应不断变化的优化需要。通过对经典函数的仿真实验,证明了该方法在提高收敛性的前提下,收敛精度较传统算法也有了提高,且克服了易陷入局部极值区域的问题。

To resolve the traditional PSO easily trapped into local optimum problem,it applies the features of chaotic motion,ergodicity,randomness and initial value sensitivity into the process of particle swarm optimization,and by the method of simulated annealing partial optimization of parameters,the parameter of particle swarm optimization algorithm changes continuously with optimization algorithm to adapt to changing optimization needs.The simulation experiments of classic function show that the method can improve the convergence,the convergence accuracy is improved,and it overcomes the problem of falling into local minimum region easily.