摘要
针对函数全局优化问题,提出了一种自适应压缩因子粒子群优化算法。研究的结果是对粒子群优化算法定义了一个与迭代步有关的压缩因子,随着迭代步不断增大压缩因子逐渐减小,使得在算法初期,压缩因子较大,提高算法的全局搜索能力,在算法后期,压缩因子较小,提高算法的局部搜索能力,另外,把差分进化算法中的交叉与变异思想引入到该粒子群优化算法中,改善了粒子的多样性。最后把算法应用到两类测试问题中,并与其他粒子群优化算法进行比较分析,数值结果表明,算法是可行的、有效的。该成果对全局优化问题的求解具有一定的参考价值和指导意义。
Particle swarm optimization algorithm with self-adaptive constriction factor is proposed for the global optimization problems. The results of research is that a new constriction factor which is related to iteration steps is defined,with the increase of iteration steps the values of the constriction factor is aregradual decrease. The values of the constriction factor are the bigger in the former process of algorithm,and the capacity of searching global optimal solution is improved. On the contrary,The values of the constriction factor are the smaller and the capacity of local search is improved in the latter process of algorithm. Moreover,the crossing and selection operation of Differential Evolution（DE） is introduced to particle swarm optimization algorithm to improve the particle diversity. In the end,the algorithm is applied to two kinds of test problems ,and numerical experiment will verify the effectiveness and feasibility of this algorithm. The resuts have a certain reference value and instructive significance to solving the global optimization problems.
出处
《辽宁工程技术大学学报（自然科学版）》
CAS
北大核心
2010年第5期949-952,共4页
Journal of Liaoning Technical University (Natural Science)
基金
国家自然科学基金资助项目(10871033)
辽宁省教育厅科学技术研究基金项目项目(2008004)
关键词
全局优化
粒子群优化算法
自适应压缩因子
global optimization
particle swarm optimization algorithm
self-adaptive constriction factor
