基于遗传交叉和多混沌策略改进的粒子群优化算法  

Improved particle swarm optimization algorithm based on genetic crossover and multi-chaotic strategies

作　　者：谭跃[1] 谭冠政[2] 邓曙光[1] Tan Yue Tan Guanzheng Deng Shuguang(School of Communication ＆ Electronic Engineering, Hunan City University, Yiyang Hunan 413000, China School of Information Sci- ence ＆ Engineering, Central South University, Changsha 410083, China)

机构地区：[1]湖南城市学院通信与电子工程学院,湖南益阳413000 [2]中南大学信息科学与工程学院,长沙410083

出　　处：《计算机应用研究》2016年第12期3643-3647,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(61471164);湖南省科技计划资助项目(2014FJ3112);湖南省教育厅优秀青年资助项目(14B033)

摘　　要：为有效改进基本PSO算法的搜索能力,提出了一种基于遗传交叉和多混沌方式改进的粒子群算法。该算法为获得比当前群体更优的最优解,采用了以下四种措施:其一,对当前群体中的最优解和每个粒子最优解进行遗传交叉操作;其二,用混沌系统动态地调整PSO算法的惯性权重;其三,对整个解空间进行混沌全局搜索;最后,对当前群体中最优解进行多维和单维的混沌局部搜索。仿真实验结果表明:与其他三种算法相比,提出的算法在解决八个整数和混合整数非线性规划问题时不仅收敛速度最快,而且具有100%的成功率。This paper proposed a particle swarm optimization （PSO） algorithm based on genetic crossover and multi-chaotic strategies to effectively improve the search abilities of the basic PSO algorithm. The proposed algorithm used four measurements to obtain the better solution than the optimal solution of the current swarm. The first was to perform a genetic crossover operation between the optimal solution of the current swarm and the best solution of each particle. The second was that a chaotic system dynamically adjusted the inertia weight. The third was a chaotic global search for the whole solution space. The last was to perform a multi-dimensional and single-dimensional chaotic local search on the optimal solution of the current swarm. Simula- tion results show that compared with the other three algorithms, the proposed algorithm not only has the fastest convergence speed, but also has a 100% success rate when solving eight integer and mixed integer nonlinear programming problems.

关 键 词：粒子群优化算法 遗传交叉 混沌惯性权重 多维和单维混沌局部搜索 混沌全局搜索 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]