自适应遗传算法解决集装箱装载问题的方法探讨  

A Discussion of Adaptive Genetic Algorithm, Solving Container-Loading Problem

作　　者：丁香乾[1] 韩运实[2] 张晓丽[2] 

机构地区：[1]中国海洋大学信息工程中心,山东青岛266071 [2]中国海洋大学计算机系,山东青岛266071

出　　处：《中国海洋大学学报（自然科学版）》2004年第5期844-848,共5页Periodical of Ocean University of China

基　　金：国家高技术研究发展计划项目 (2 0 0 2AA41 4 0 1 0 )资助

摘　　要：集装箱装箱问题是 1个有很强应用背景的组合优化问题 ,约束条件多 ,求解极为困难。本文探讨了自适应遗传算法在复杂集装箱装载问题中的应用 ,算法中采用跨世纪精英选择策略保持了群体的多样性 ,并给出了有效的解码算法。实例仿真结果显示出很好的效果 ,同等条件下求解结果比使用启发式算法高出近 10个百分点 ,表明该算法是行之有效的。Container loading problem is a combinatorial optimization problem with a broad application background. It involves constraint complexity, so it is difficult to obtain an optimal solution. In this paper, the adaptive genetic algorithm for complex container-loading problem is studied and an effective decoding algorithm is presented. Cross generational elitist selection, which aims at keeping the population's diversity, is employed in the algorithm. Simulation result, which is rational, is 10 percent higher than that of the heuristic algorithm solution under an opposite situation, so it demonstrates that the genetic algorithm is an effective approach to solve a complex container loading problem.

关 键 词：遗传算法 集装箱装载 自适应 空间划分 组合优化 

分 类 号：TB114[理学—概率论与数理统计] TP39[理学—数学]