一种改进遗传算子的遗传算法组卷策略  

An improved genetic algorithm test paper generation strategy based on genetic algorithms

作　　者：张净宇 王靖 吴志雄 王旭 丁宇[1] ZHANG Jingyu;WANG Jing;WU Zhixiong;WANG Xu;DING Yu(School of Computer Science and Technology,Yangtze University,Jingzhou Hubei 434020,China)

机构地区：[1]长江大学计算机科学学院,湖北荆州434020

出　　处：《智能计算机与应用》2023年第5期40-45,共6页Intelligent Computer and Applications

摘　　要：针对计算机辅助教学中的智能组卷问题,本文通过建立多约束条件下组合优化的数学模型,提出了一种改进遗传算子的遗传算法,采用最优个体保存策略与轮盘赌算法结合方法选择算子;交叉和变异算子采用随种群进化过程中个体的适应度的变化而自适应调整的种群交叉与变异概率;采用分段十进制编码方式以提高算法性能。改进遗传算子的遗传算法种群收敛速度更快,并在一定程度上避免了算法陷于局部最优。将改进遗传算子的遗传算法应用于组卷系统中,能更快地生成试卷且质量能满足用户需要。Aiming at the problem of intelligent test paper generation in computer-aided teaching,this paper proposes a genetic algorithm to improve the genetic operator by establishing a mathematical model of combinatorial optimization under multi-constraint conditions,compared with the traditional genetic algorithm,the genetic algorithm of the improved genetic operator adopts the combination method of optimal individual preservation strategy and roulette algorithm,and the crossover and mutation operator adopts the population crossing and mutation probability that adaptively adjusts with the change of individual fitness in the process of population evolution.At the same time,the piecewise decimal encoding method is adopted to improve the algorithm performance;The population convergence speed of the genetic algorithm of the improved genetic operator is faster,and in addition,the algorithm is avoided from falling into local optimization to a certain extent.Finally,the genetic algorithm of improved genetic operator is applied to the group volume system,and it is found that the proposed strategy can generate test papers faster and the quality can better meet the needs of users,which reflects the superiority of the improved algorithm.

关 键 词：多约束条件 改进遗传算子 遗传算法 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]