摘要
在鱼群算法优化的研究中,针对人工鱼群算法(AFSA)存在的速度慢、精度差、早熟收敛等问题,提出一种新的改进人工鱼群算法,即一种采用动态游动模式的鱼群算法(DSMFSA)。上述算法让每条"鱼"具有多种搜索模式,让每条"鱼"具有机动搜索食物的能力,并可根据群体信息的反馈和自身状态随时调整搜索方式。在数值实验中选择了几个比较典型的基准函数,用来测试上述算法的性能。实验结果表明:DSMFSA算法大大改善了人工鱼群算法(AFSA)存在的易陷入局部最优、优化精度不高之不足,明显具有比AFSA好得多的优化性能。说明改进算法具有跳出局部最优的能力,可用于求解高维的复杂优化问题。
In order to overcome the shortcoming of artificial fish swarm algorithm （AFSA）, a new improved AF- SA, called a fish swarm algorithm by using dynamic swimming- modes （DSMFSA）, is proposed in this paper. The approaches of the DSMFSA are that make each fish have multi - search - modes and have the ability of dynamic searching food, and make each fish, at any time, can adjust its search - modes according to the feedback of the swarm - information and itself state in its search process. In order to test the performance of the DSMFSA, experiments are done on some typical benchmark functions in our numerical experiments, The experimental results show that the DSMFSA can overcome the shortcoming of the AFSA of easy being trapped into the local optimal and the defect of the low accuracy of optimization, and has better optimization performance than that of the AFSA. It indicates that DSMFSA has the ability of jumping out of local optimal, and can be used to solve the high dimensional and the complex optimization problems.
出处
《计算机仿真》
CSCD
北大核心
2015年第4期208-215,共8页
Computer Simulation
基金
广西自然科学基金(0832084)
广西高等学校科研项目(KY2015YB078)
