摘要
提出一种含全交叉算子的遗传算法。与传统的基于点的交叉不同,全交叉算子选择与、或、异或3种方式中的一种作为染色体之间的交叉方式。为了提高算法进化速度,从种群中选择优秀个体组成精英集,每轮更新精英集,种群中染色体进行交叉操作时会先选择一种交叉方式,再从精英集中随机选择精英个体作为交叉染色体进行交配。在同样情况下,与单点交叉、两点、多点交叉进行比较,仿真实验结果表明,含全交叉的改进遗传算法有较好的优化效果。
Differing from the traditional crossover operator, a genetic algorithm containing a to-tal crossover operator is proposed in this paper, which uses A N D , OR, XOR operators as a cross method. To improve evolution speed of the algorithm, the algorithm selects elite individuals from the population to form an elite set, and the elite set updates per round. A cross method and an elite individual from the elite set are selected when chromosomes conduct crossover. Simulations show that the improved genetic algorithm w ith total crossover has better optimization effect as compared with single point crossing, two-point and multi-point algorithms.
出处
《上海电机学院学报》
2017年第4期196-200,214,共6页
Journal of Shanghai Dianji University
基金
国家自然科学基金项目资助(61374136)
上海市人才发展基金项目资助(201511)
关键词
遗传算法
全交叉算子
交叉方式
精英集
genetic algorithm (G A)
total crossover operator
cross method
elite set