基于特征进化的混合遗传算法

HYBRID GENETIC ALGORITHM BASED ON CHARACTERISTIC EVOLUTION

对于遗传算法存在早熟性收敛和收敛速度慢等问题,可通过保护存在于种群中的最小诱导模式和属于收敛优化解或全局最优解的有效基因块,得到有效的改善。通过对种群中个体之间关系分析,建立特征保护策略及特征进化算子,由此改进的混合遗传算法具有较高的收敛速度,并能收敛于规模小于2 000个城市的旅行商问题全局最优解。

The two problems in genetic algorithm, premature convergence and slow-footed convergence, can be efficiently modified by protecting the minimum inducing schema and the effective gene blocks belonged to convergence optimal solution or global optimal solutions, both of them exist in population. Characteristic protection policy and characteristic evolution operators are put forward through analyzing the relationship among individuals in population. New hybrid genetic algorithm improved with them has more fast convergence speed, and is able to converge to global optimal solution of TSP whose scale is less than 2,000 cities.