求解强异类集装箱装载问题的混合蚁群算法  

Hybrid binary ant colony algorithm for strongly heterogeneous container loading problem

作　　者：魏平[1] 熊伟清[1] 

机构地区：[1]宁波大学电子商务与物流研究所,浙江宁波315211

出　　处：《计算机工程与应用》2013年第7期252-257,共6页Computer Engineering and Applications

基　　金：浙江省自然科学基金(No.Y1100052);浙江省教育厅科研项目(No.Y201017000)

摘　　要：针对强异类集装箱装载问题,设计了一种混合蚁群算法。算法中搜索空间分为货物摆放的优先序列和货物摆放的状态两部分;引入体积大的货物优先放入的启发式规则;将蚂蚁搜索得到的序列与历史最优序列进行交叉,取三者最优序列作为该蚂蚁的搜索路径;在更新信息素时,采取两种挥发系数更新信息素以避免信息素过快饱和,同时分析了算法的复杂度。通过三个强异类实例的测试,表明算法得到的装载方案有较高的空间利用率。Aiming at the strongly heterogeneous Container Loading Problem （CLP）, a mixed Ant Colony Algorithm （ACO） is designed. The solution of problem is divided in two parts, the priority of the goods and the goods＇ state. Based on heuristic rules, the larger goods have priority to pack in container, so volume is considered as heuristic information. The sequence that ant has searched crosses with historical optimal sequence. The optimal one among the three sequences is choose as the wanted sequence. In order to avoid pheromone over-rapid saturated, pheromone is updated by adopting two volatile coefficients. The complexity of the algorithm is analyzed. Through testing three examples, the space utilization is high by using this algorithm.

关 键 词：集装箱装载 蚁群优化算法 启发式规则 整数规划 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]