遗传算法的改进及其在排课问题中的应用  

Improvement of Genetic Algorithm and Its Application in Course Scheduling

作　　者：姜婧 白似雪[1] JIANG Jing;BAI Sixue(College of Information Engineering,Nanchang University,Nanchang 330031,China)

机构地区：[1]南昌大学信息工程学院,江西南昌330031

出　　处：《南昌大学学报（理科版）》2018年第4期388-392,共5页Journal of Nanchang University(Natural Science)

基　　金：江西省科技厅重大科技攻关项目(20152ACE50021);江西省科技成果重点项目(20151BBI90008);2013年度"洪城计划"人才项目13号

摘　　要：分析基本遗传算法中传统选择、交叉策略的不足,结合多种选择方法及模拟退火机制,提出改进的选择策略;在交叉阶段,利用蚁群算法的正反馈机制改进了交叉策略;采用最大允许停滞代数的方法来结束算法的运行。由在排课问题中的对比实验可以看出,本文提出的改进算法改善了遗传算法局部搜索能力弱的缺点,提高了算法运行效率。The shortcomings of traditional selection and cross strategies in basic genetic algorithms are investigated in this paper.Combined with a variety of selection methods and simulated annealing mechanisms,we propose an improved selection strategy,where the crossover strategy in the cross stage is improved by using the positive feedback mechanism of the ant colony algorithm.The maximum permissive stagnation algebra is adopted to design the stopping criterion of the algorithm.As can be seen from the comparison experiment in the course scheduling problem,the proposed algorithm improves the local searching ability and the efficiency of the genetic algorithm.

关 键 词：选择 交叉 模拟退火机制 正反馈机制 排课问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]