改进的人工鱼群混合算法在智能组卷中的应用  

Improved AFSA for Solving Intelligent Test Problem

作　　者：吴旭[1] 洪联系[1] 魏德志[1] 郭剑平[1] 

机构地区：[1]集美大学诚毅学院,福建厦门361021

出　　处：《集美大学学报（自然科学版）》2012年第5期394-400,共7页Journal of Jimei University：Natural Science

基　　金：集美大学诚毅学院重点教学改革基金资助项目(P10222)

摘　　要：现有的智能组卷多采用单一算法,而每种算法都有其各自的缺点,针对此缺陷提出了结合人工鱼群算法和遗传算法的优点组成混合智能组卷算法.在智能组卷开始时,采用人工鱼群算法快速靠近组卷目标,在组卷过程中,当最优个体在连续多个迭代过程中无变化或变化极小时采用遗传算法对人工鱼个体进行跳变,提高收敛速度.通过模拟计算证明,该混合智能算法能有效地优化其中单一算法独自进行智能组卷的成效.Intelligent auto-generating paper was a optimization problem which had multi-target parameter in certain constraint conditions.There had been many algorithms to realize intelligent auto-generating paper.But,most of the methods achieving the target used only a single algorithm.Since each of the algorithm had its own shortcomings,so in the process of intelligent auto-generating paper,it sometimes can＇t avoid the defects of a single algorithm.By incorporating AFSA and GA＇s advantages into a mathematical model for intelligent auto-generating paper,a HIDA algorithm was proposed.In the course of generating paper,the AFSA was adopted to approach to the target first.When the best individual didn＇t change or changed very small continually,the GA was used to make the fish jump to increase the convergence speed.The results of experiments showed that the HIOA-GFA was better than single one of the Algorithms to generate paper.

关 键 词：智能组卷 人工鱼群 遗传算法 混合智能算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]