基于改进遗传算法的智能组卷研究  

On Intelligent Test Paper Auto-Generation Based on Improved Genetic Algorithm

作　　者：陈晓敏[1] 梁静[2] 葛宇[3] 

机构地区：[1]成都电子机械高等专科学校信息与计算科学系,成都610031 [2]成都电子机械高等专科学校网络中心,成都610031 [3]四川师范大学基础教学学院,成都610068

出　　处：《西南师范大学学报（自然科学版）》2012年第5期98-101,共4页Journal of Southwest China Normal University(Natural Science Edition)

摘　　要：通过在遗传算法中引入个体浓度的选择机制和记忆机制,确保了进化过程中种群内个体的多样性,避免局部收敛,保证了算法朝优化方向进化.实验结果表明改进算法能跳出局部收敛,有效避免了早熟产生和遗传退化现象出现.Intelligent test paper auto-generation deals with a multi-objective combinatorial optimization. As a common tool to generate test paper automatically, Genetic algorithm could not achieve the best solution frequently due to the local convergence. In this paper the concentration mechanism and memory mechanism are combined into the genetic algorithm to ensure the population diversity and avoid local convergence. Experiment results show the proposed algorithm can be applied to generate test paper automatically and effectively.

关 键 词：组合优化 智能组卷 遗传算法 局部收敛 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]