基于改进遗传算法的智能组卷研究  

RESEARCH ON AN INTELLIGENT TEST CONSTRUCTION METHOD BASED ON IMPROVED GENETIC ALGORITHM

作　　者：王萌[1] 唐培和[1] 李春贵[1] 

机构地区：[1]广西工学院计算机工程系,广西柳州545006

出　　处：《计算机应用与软件》2008年第2期49-51,共3页Computer Applications and Software

基　　金：广西科学基金资助项目(桂科自0481016;桂科自0640034)。

摘　　要：组卷问题是一个多目标约束优化问题,采用集合论思想,将智能组卷的要求形式化,提出了一种基于改进遗传算法来进行智能组卷的方法。该算法针对传统智能算法组卷速度慢、成功率低的缺点,针对简单遗传算法的编码、遗传算子存在的问题进行了改进。实验结果表明,改进的算法在保证试题产生的质量前提下,提高了智能组卷的运行速度,具有较好的实用性。Computer test construction is a multi-objective constrained optimization problem. Set principle is applied to formalized the requirement of computer test construction, and a new intelligent test construction method based on improved genetic algorithm is proposed. The algorithm has been improved in the aspects of convergence, success rate, codes and genetic operators. The experiments show that the improved genetic algorithm not only enhances the quality of test construction, but also reduces the time of execution, and it has strong practicability.

关 键 词：改进遗传算法 智能组卷 编码 遗传算子 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]