基于信息熵的粒子群优化算法  

Improved Particle Swarm Optimization Algorithm Based on Information Entropy

作　　者：姜毅[1] 乐庆玲[2] 

机构地区：[1]武汉科技大学计算机学院,武汉430081 [2]湖北大学档案系,武汉430062

出　　处：《科学技术与工程》2008年第9期2352-2355,共4页Science Technology and Engineering

基　　金：国家自然科学基金(O60705012)资助

摘　　要：粒子群算法是一种进化计算技术,成功地运用于广泛的数值优化问题。PSO算法在求解高维复杂函数优化问题时容易陷入局部最优。有鉴于此,提出了一种基于信息熵的粒子优化算法。该算法提高设计了一种兼顾种群选择性压力以及种群多样性的选择策略,从而提高了粒子在运行过程中的多样性。实验表明,该算法有效避免了陷入局部最优,提高了全局最优解的搜索精度。Particle Swarm Optimization （PSO） algorithm which has been shown to successfully optimize a wide range of continuous functions. PSO easily plunge into the local minimum when it solve the complexhigh dimension function optimization problem. Thus a new PSO based on information entropy is proposed, which given attention to population selectived pressure and population diversity, improve the diversity of swam. The algorithm can not only escape from local minimum, but also enhance the capability to search the global optimization.

关 键 词：粒子群优化算法 信息熵 种群多样性 

分 类 号：TP301.9[自动化与计算机技术—计算机系统结构]