摘要
提出了一种引入个体相异度阀值函数的新自适应遗传算法,该算法根据个体的相异性,给出了个体相异度的概念和相应的计算公式,并设计了一个与进化代数相关的阀值函数,以实现选择性交叉和变异.同时为了克服传统自适应遗传算法在进化过程中停滞不前的缺点,该算法引入非线性函数作为自适应交叉率和变异率计算公式.最后,针对典型车间调度问题,分别对改进算法和其他优化算法的计算结果进行了比较,结果表明新算法更有效.
An improved adaptive genetic algorithm with individual diversity threshold function was proposed. The concept of individual diversity and relevant calculating formula based on the individual diversity was presented, and the threshold function relevant to evolutional algebra was devised to realize the selectivity intersection and variation. Meanwhile, a nonlinear function as the calculating formula of adaptive intersection and variation rate was drawn into the improved algorithm to overcome the disadvantages of traditional adaptive genetic algorithm which is liable to bog down in the evolutional process. Finally, the calculation result of the improved algorithm was compared with other optimal algorithms in solving classic job-shop scheduling problems.
出处
《大连交通大学学报》
CAS
2009年第4期60-63,共4页
Journal of Dalian Jiaotong University
关键词
相异度阀值
个体相异度
最优保存
自适应
diversity threshold
individual diversity
elitist preserved
self-adaptive
