psoparticleswamoptimization粒子群优化算法应用-read.doc

目录 言 1 第一章 概述 2 1.1 引言 2 1.2 研究背景 2 1.2.1 人工生命计算 2 1.2.2 粒子群算法与遗传算法 3 1.2.3 人工神经网络与粒子群优化算法 3 第二章 粒子群优化算法 5 2.1 基本粒子群优化算法 5 2.2 算法流程 5 2.3 基本粒子群优化算法的缺点及改进方法 6 2.3.1 基本粒子群优化算法的缺点 6 2.3.2 几种改进算法 6 标准粒子群优化算法 6 带有收缩因子的粒子群优化算法 7 采用微分扰动的粒子群 8 带有小生境拓扑结构的粒子群优化算法 9 带有梯度加速因子的粒子群优化算法 10 第三章 基于多种群的粒子群优化算法 11 3.1 算法基本改进方法 11 3.2 算法伪代码 11 第四章 实验分析 14 4.1 几种常用改进算法 14 4.2 标准测试函数 14 4.3 收敛速度实验分析 15 4.3.1 实验设置 16 4.3.2 实验结果 16 4.4 收敛性能实验分析 17 4.4.1 实验设置 17 4.4.2 实验结果 18 4.5 算法参数分析 19 4.5.1 实验设置 19 4.5.2 实验结果 20 第五章 基于多种群的粒子群优化算法的应用 21 5.1 生产计划问题描述 21 5.2 实例 22 5.3 运行结果 23 第六章 总结与展望 24 6.1 课题总结 24 6.2 后续研究展望 24 参考文献 25 致谢 27  摘 要 在智能领域，大部分问题都可以归结为优化问题。常用的经典优化算法都对问题有一定的约束条件，如要求优化函数可微等，仿生算法是一种模拟生物智能行为的优化算法，由于其几乎不存在对问题的约束，因此，得到广泛应用。本 Abstract In the field of artificial intelligence, most problems belong to the category of optimization. The classical algorithms in common use are usually limited by certain optimizing problems such as the object optimization function, which has to be a continuous one. As bionic algorithm imitates the intellective actions of life free from the limits resulting from the optimizing problems, this kind of algorithm is commonly used. At first place, this dissertation, basing on the particle swarm optimization, illustrates the fundament of this algorithm. This paper also places emphasis on the analysis of parameter that may affect the performance of the algorithm. Also an introduction of improved algorithms and popular advanced improving strategies is shown, as well as the mastery of basic researching process and methods. According to the result of the analysis, the author put forward a new multi-swarms PSO algorithm to overcome the defects of the original. Through the simulation, the results show that, compared with other PSO variants, the algorithm proposed by the author has attained a better solution to the same problems. Finally, the paper gave some further expectations concerning the PSO algorithm research. Keywords: Pa