融合powell搜索法的粒子群优化算法-控制与决策.pdf

第27 卷 第3 期 控 制 与 决 策 2012 年 3 月 Vol. 27 No. 3 Control and Decision Mar. 2012 文章编号: 1001-0920 (2012) 03-0343-06 融合Powell 搜索法的粒子群优化算法 a,b a b 吴建辉 , 章 兢 , 陈红安 (湖南大学 a. 电气与信息工程学院，b. 信息科学与工程学院，长沙410082) 摘 要: 为了进一步提高多模态函数寻优的效率, 提出一种融合Powell 搜索法的粒子群优化算法. 将PSO 算法的全 局搜索能力与Powell 法的强局部寻优能力有机地结合起来, 在保证求解速度, 尽可能找到全部极值点的同时提高了 解的精确性. 由于该算法只利用了函数值信息而不需要计算导数, 是求解可微和不可微多模态函数优化问题的通用 方法. 仿真实验表明了新混合算法的有效性. 关键词: 多模态函数；Powell 搜索法；粒子群优化 中图分类号: TP18 文献标识码: A Particle swarm optimization algorithm combination with Powell search method a,b a b WU Jian-hui , ZHANG Jing , CHEN Hong-an (a. College of Electrical and Information Engineering，b. College of Information Science and Engineering ，Hu’nan University ，Changsha 410082，China ．Correspondent ：WU Jian-hui ，E-mail ：wujianhui123@) Abstract: In order to further improve the efficiency of multi-modal function optimization, a novel hybrid algorithm which combines improved particle swarm optimization algorithm and Powell search method is proposed. The Powell-IPSO hybrid algorithm organically integrates PSO algorithm which has powerful global search capability with Powell search method which has strong local search ability. The hybrid algorithm ensures quick convergent speed and find all extreme points as much as possible, and solution’s precision is improved. The hybrid algorithm only uses function values information and doesn’t need to calculate the derivative of function. So it is a generalized algorithm for multi-modal function of differentiable and non-differentiable. The simulation experionents sh