【摘要】 粒子群优化算法是一种基于群体智能的随机优化算法,针对粒子群优化算法稳定性较差和易陷入局部收敛的缺点,作者提出了一种基于tent映射和logistic映射的粒子群算法,一方面,分别应用tent映射和logistic映射初始化均匀分布的粒群提高了初始解的质量;另一方面,设定粒子聚集程度的量化计算公式和判定阈值,并引入自适应高斯变异策略,增强了算法跳出局部最优解的能力.通过对基准测试函数的仿真计算,证明该算法具有稳定性好和收敛速度快的特点.更多还原

【Abstract】 Particle swarm optimization(PSO) is a population-based stochastic optimization originating from artificial life and evolutionary computation.PSO,however,has a feature of un-stability during its running,and like other evolutionary algorithms,has a tendency to get stuck in local optimal solutions during the search process.So,two improved particle swarm optimization are proposed in this paper,which are PSO with an initial population of tent map solutions and Gaussian mutation based on maximal focus distance(Tent-PSO) and PSO with an initial population of logistic map solutions and Gaussian mutation based on maximal focus distance(Logistic-PSO),respectively.Simulation results on two benchmark functions illustrate that the PSO proposed in this paper is feasible and efficient.更多还原