教学资源配置优化中遗传算法的应用与改进  

Application and Improvement of Genetic Algorithm for Optimization in Allocating Teaching Resources

作　　者：严宏[1,2] 

机构地区：[1]中国民用航空飞行学院,四川广汉618307 [2]四川大学视觉合成图形图像技术国家重点学科实验室,四川成都610064

出　　处：《计算机技术与发展》2016年第3期130-134,139,共6页Computer Technology and Development

基　　金：中国民用航空局民航科技创新引导资金(C2013041)

摘　　要：对于临时新增教学任务,教学资源安排不同于学期开始前的教学资源配置,通常具有特定的配置需求。在教学资源有限的情况下,为了使新增教学班级配置的教室更加一致或接近,避免频繁变动教室给教学带来的不便,使用遗传算法进行教室资源的分配,并根据应用需求进行了改进。其中设计了更加适宜的十进制编码,提出的适应度函数能有效地对教室资源的一致或接近程度进行量化计算。根据染色体编码特点和优化目标,对遗传算子进行选择和改进,特别对单点交叉进行改进尽可能地提高交叉后的适应度,同时保持种群的多样性。实验结果表明,改进的遗传算法相对于标准遗传算法在用时更少的情况下得到的教学资源分配更加满足优化目标,计算效率得以提高,并且成功应用于实际的教学资源配置。For temporarily added curriculum, the arrangement of teaching resources which usually has some specific requirements is different from that done before a semester begins. In the case of limited teaching resources, an improved genetic algorithm was proposed for optimizing the consistency and nearness of the classrooms allocated for temporary curriculum, with aim to avoid the inconvenience to teach- ing caused by frequent change of classrooms. Based on the requirement of the given application in this paper,a more efficient decimal encoding was developed. Correspondingly, a suitable fitness function was proposed to quantitatively calculate the consistency and nearness of the allocated classrooms. According to the characteristic of the chromosome encoding and the optimization goal, the algorithm selected existing genetic operators and improved them, especially the modified one-point crossover that attempts to enhance the fitness value and maintain the diversity of population. The experimental results show that the improved genetic algorithm, which has been successfully applied, can achieve more optimal solution within a shorter period of time when compared with the standard genetic algorithm.

关 键 词：资源配置 遗传算法 十进制编码 适应度函数 交叉算子 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]