基于增强个体信息交流的蜻蜓算法

摘要

提出一种基于增强个体信息交流的蜻蜓算法(EIDA).针对基本蜻蜓算法存在的不足,EIDA引入三种策略:贪婪策略、平衡策略和组合策略.贪婪策略的核心在于充分利用上下代优秀个体的信息,提高收敛速度;平衡策略用于个体从全局搜索向局部开发过渡,使算法更稳定;组合策略则用于增强局部开发能力,提高解的精度.八个典型复杂函数的仿真测试表明,EIDA的全局搜索能力强,收敛速度快,求解精度高.%A new dragonfly algorithm based on enhancing the exchange of individuals'information(EIDA)is proposed. In view of the shortcomings of dragonfly algorithm, EIDA uses three strategies:a greedy strategy, an equilibrium strategy and a combined strategy. Greedy strategy aims at fully using the best individuals'information between the parent and child generation, in order to speed up the convergence. Equilibrium strategy is used for balancing exploration and exploita-tion, and stabilizing the algorithm. Combined strategy is also applied, with the purpose of enhancing exploitation and improve the precision of solution. Experimental results on eight representative complicated functions indicate that EIDA has powerful exploration, fast convergence and high precision.