一类求解箱式约束优化问题的自适应引力搜索算法

Self-adaptive Gravitational Search Algorithm for Box-constrained Optimization Problems

为了改进引力搜索算法求解箱式约束优化问题的性能,提出了一类自适应引力搜索算法,新算法定义了算法停滞系数,当算法陷入停滞时,可以自适应的修改引力参数,帮助算法跳出停滞状态;定义了个体相似系数,当种群陷入局部最优时,通过变异策略改善种群的多样性;数值试验结果表明,新算法有效的平衡了全局开发和局部搜索能力,具有更强的全局寻优能力,适于求解复杂优化问题。

To improve the performance of Gravitational Search Algorithm （GSA） for box-constrained optimization problems, an im proved algorithm based on self-adaptive gravitational search algorithm was proposed. Stagnation coefficient and similarity coefficient are defined. When algorithm has been in stagnation behavior, gravitation parameter is revised adaptively to iump out of stage of stagnation. When swarm fall into local optimal, diversity of swarm will be improved by mutation strategy. The numerical experiment on benchmark functions shows that the improve algorithm efficiently balances the exploit and explorer, especially suitable for solving high dimension and multimodal function optimization problem.