基于惯性因子的混沌粒子群优化算法研究

粒子群优化算法本身在多峰复杂函数时会出现早熟收敛现象,降低粒子的多样性,导致粒子群不能收敛到全局极值点。针对粒子群优化算法的局限性,把混沌优化思想引入到粒子群算法,采用混沌优化粒子群算法对测试函数进行仿真,并在此基础上加入惯性因子对混沌优化粒子群算法进一步改进,Matlab仿真结果表明,改进的混沌优化粒子群算法,结合了混沌和粒子群算法共同的优点,能快速、准确地搜索到全局最优值。

Research on chaos particle swarm optimization algorithm based on inertia weight

When particle swarm optimization（PSO） algorithm is used in complicated multi-modal functions,premature convergence will occur,which can reduce the diversity of particles and cause the particle swarm algorithm unable to converge to the value of global optimal solution.To overcome the defect of premature convergence on PSO algorithm,the idea of chaos is introduced into PSO algorithm called chaos optimization particle swarm algorithm（CPSO）.In this paper,CPSO is employed to simulate test functions,on the basis of which an inertia factor is introduced to improve the performance of CPSO.Matlab simulation results show that compared with the original particle algorithm,the improved CPSO algorithm has the advantages of both the chaos and particle swarm optimization（PSO） algorithm and can obtain the global optimum value quickly and accurately.