【摘要】 提出一种基于优胜劣汰规则的异类多种群蚁群算法,该算法由多类不同特性的蚁群构成,彼此间具有潜在的合作性和对抗性。根据蚁群间定期信息交换的结果,引入自然界优胜劣汰准则,设定蚁群间的合作规则、竞争规则、裂变规则。以旅行商问题为例进行相关实验和比较。通过多个种群间的相互合作与竞争,保留优势种群,淘汰劣势种群,提高求解效率,改善解的多样性,使算法更容易收敛到全局最优解。更多还原

【Abstract】 A Heterogeneous Multiple Ant Colony algorithm based on Survival of Fittest rules(HMACSF) is presented.This algorithm introduces more than one type of ant colony.All types of ant colonies with different pheromone updating mechanism and searching traits have mutual compensation of advantages,as well as mutual competitive exclusion.According to the results of the exchanging,HMACSF retains the dominant colonies,weeds out the inferiors,and improves the solving efficiency and diversity of solutions,to easily converge the global optimal solution.A series of Traveling Salesman Problem(TSP) experiments show that this algorithm can generate solutions with better quality and faster speed.更多还原