带变异算子的双种群粒子群优化算法

Novel bi-group particle swarm optimization with mutation operator

提出一种带变异算子的双种群粒子群算法,搜索在两个不同的子群中并行运行,分别使用不同的惯性权值,使得种群在全局和局部都有较好的搜索能力。通过子群重组实现种群间的信息交换。在算法中引入变异算子,产生局部最优解的邻域点,帮助惰性粒子逃离束缚,寻得更优解。对经典函数的测试结果表明,改进的算法在收敛速度和精度上有更好的性能。

An algorithm ofbi-group particle swarm optimization with mutation operator is proposed, by searching the two sub-groups which are parallel performed and have inertia weights separately. A better searching ability in both partial and overall situations is realized. By exchanging information in sub-groups that are reorganized and adopting mutation operator, a neighborhood spot in partial optimal solution will be found, and this will help the inert operator flee away from restrictions for a superior solution. The testes of classical functions show that, this improved algorithm has a better performance in both the convergence rate and the precision.