《用混沌粒子群算法求解函数优化问题》-毕业设计（论文）.docVIP

PAGE lw 本科生毕业设计报告 课 题： 用混沌粒子群算法求解函数优化问题 作 者： 专 业： 电子信息科学与技术 班 级： 信科80802班 指导教师： 2012年 5 月 9日 lw Yangzhou University Graduation Design (Thesis) A Chaos Particle Swarm Optimization Algorithm On Function By Dai Yue Supervised by Lecturer Zhu hai mei MAY, 2012 摘 要 粒子群在搜索过程中容易陷入局部而无法找到全局最优值，且算法后期的粒子速度下降过快而失去搜索能力等缺陷，为了解决此早熟问题，提出了一种基于混沌思想的新型粒子群算法。该算法首先通过混沌方法初始化粒子的初始位置和速度，增强了粒子的搜索能力。算法还通过混沌序列得到的惯性权重取代传统的线性递减的惯性权重，使粒子速度呈现多样性的特点，从而提高算法的全局搜索能力；对比仿真实验表明，本文所提出的混沌粒子群算法较传统粒子群算法具有更好的搜索性能。 本文是基于VC++6.0平台，通过C语言实现基本粒子群算法和改进的混沌粒子群算法。通过几个经典测试函数进行实验，列出实验结果。仿真实验结果表明：混沌粒子群算法比基本粒子群算法具有更好的收敛性能和寻优能力。 关键词：粒子群算法；混沌粒子群；惯性权重 ABSTRACT Particle swarm optimization was easily trapped into the local optima and failed to find the global optima. In the late period, it is easy to lose search ability because of the fast particle velocity decrease. To solve this premature problem, a new thought based on chaotic particle swarm optimization algorithm is proposed in this paper. In the improved algorithm, method of chaos is used to initialize the position and speed of particles, which enhances the search ability of particles. Furtherore, self-adapting inertia generated from the chaotic sequence replaces the traditional linear decreasing weight of inertia, which results in the diversity of particles’ speed ,so as to improve the the global search ability. Simulation results demonstrate that the presented chaotic particle swarm algorithm has better search performance than the traditional particle swarm algorithm. Based on vc + + 6.0 platform and C language, basic particle swarm algorithm and improved chaotic particle swarm algorithm are simulated. A few classic test functions are tested, and the experimental results are listed. The simulation results show that: chaotic particle swarm algorithm has better convergence and optimization ability than the basic PSO. Key words：Particle Swarm Optimization ; Chaotic Particle Swarm ; Inertia Weight 目 录 1 绪论 1