摘要
利用正交实验法的全局均衡思想,提出了一种采用多点正交交换的遗传算法。算法通过正交表安排遗传算法的交换运算,并在所产生的多个子代中选择适应度大的进入下一次进化,这样既加快了算法的收敛速度又保证了种群的多样性。实验证明,该算法不但可以有效地克服标准遗传算法的缺陷,而且计算速度、精度和算法稳定性也得到了显著提高。
Using the global equilibrium design ideology of orthogonal experiment method, this paper proposes the genetic algorithm with multi-point orthogonal crossover operation. Crossover operation of the algorithm is based on the orthogonal array, and the two of many offspring that have bigger fitness are chosen to put in next evolution. The algorithm can ensure population multiformity and convergence speed rapidly. The research results show that the algorithm can not only overcome the short comings of SGA effectively, but also evidently improve the computing speed, computing precision and stability.
出处
《计算机工程》
EI
CAS
CSCD
北大核心
2005年第24期151-152,158,共3页
Computer Engineering
基金
国家"十五"
"211"一期学科建设基金资助项目"信息安全保密技术与相关数学理论研究"
关键词
遗传算法
正交试验
多点交叉
仿真
Genetic algorithm
Orthogonal experiment
Multi-point crossover
Simulation
