基于改进遗传算法的快速自动组卷算法研究  

New algorithm for intelligent test paper composition based on improved genetic algorithm

作　　者：陈国彬[1] 张广泉[2,3] 

机构地区：[1]重庆工商大学融智学院,重庆400033 [2]苏州大学计算机科学与技术学院,江苏苏州215006 [3]中国科学院计算机科学国家重点实验室,北京100080

出　　处：《计算机应用研究》2015年第10期2996-2998,3003,共4页Application Research of Computers

基　　金：重庆市教委科学技术研究项目(KJ133103);江苏省自然科学基金资助项目(BK2011152);中国科学院计算机科学国家重点实验室开放课题(CSYSKF0908)

摘　　要：为了克服现有的试题管理系统在试卷生成环节的速度和质量上存在的缺陷,将粗粒度并行遗传算法与自适应技术相结合,提出了一种自适应调整种群迁移的快速并行遗传算法。分别从试题库的编码方案、遗传策略、适应度函数的优化、交叉变异算子的选择和自适应度值函数的选取等方面进行设计,取得了很好的适应度,同时采用并行的策略,提高了算法的运行速度。仿真实验结果表明,该算法能够成功应用于自动组卷,并且组卷效率和成功率都得到了明显的提高,具有很好的通用性。该算法用于智能快速自动组卷是可行的、有效的。In order to overcome the existing exam management system on the speed and quality of test paper generating links exist defects,the coarse-grained parallel genetic algorithm combined with the adaptive technology,this paper proposed a fast parallel genetic algorithm of adaptive adjustment of population migration. Respectively from the examination of coding scheme,the optimization of genetic strategy,fitness function,crossover mutation operator selection and the selection of fitness value function,etc to carry on the design,had obtained the very good fitness,adopted the tactics of parallel at the same time,improved the running speed of the algorithm. Simulation experiment results show that the algorithm is successfully applied to the automatic group volume,volume and group to improve the efficiency and success rate are obvious,has the very good generality. The algorithm used in the intelligent fast automatic group volume is feasible and effective.

关 键 词：试题管理系统 数据挖掘 组卷算法 遗传算法 粗粒度 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]