摘要
为了寻找复杂多峰函数的全局最优解,在标准粒子群优化算法的基础上,提出一种基于不同行为的两分群交换粒子群优化算法。该算法将微粒分成大小相同的2个种群,不同种群采用不同进化模型。利用不同进化模型具有不同进化行为的特点,两分群相互影响并促进。该方法可以保持种群多样性,降低陷入局部极值的可能性。对一些复杂函数的仿真结果表明,该算法易于找到全局最优解。
In order to locate the global optimum of complex multimodal function, on the basis of standard Particle Swarm Optimization(PSO) algorithm, this paper proposes a two sub-swarms exchange PSO algorithm based on different behaviors. The particles are divided into two swarms in the same size. Different swarms adapt different evolution models. By using the characteristics that different evolution models have different evolution behaviors, the two sub-swarms influence and promote each other. This method can maintain diversity of population and reduce the possibility of local minimum, Simulation results of some complex functions show that the algorithm can easily find the global optimum solution.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第7期176-178,共3页
Computer Engineering
基金
国家自然科学基金资助项目(50539020)
江西省自然科学基金资助项目(2007GZS1056)
江西省教育厅科技基金资助项目(GJJ10630)
关键词
粒子群优化
种群多样性
全局最优解
Particle Swarm Optimization(PSO)
diversity of population
global optimum solution