运筹与管理
運籌與管理
운주여관리
OPERATIONS RESEARCH AND MANAGEMENT SCIENCE
2015年
2期
51-57
,共7页
运筹学与控制论%冲突装箱问题%最大团%启发式算法
運籌學與控製論%遲突裝箱問題%最大糰%啟髮式算法
운주학여공제론%충돌장상문제%최대단%계발식산법
operations research and cybernetics%bin packing problem with conflicts%maximum clique%heuristic algorithm
现实物流活动中大量存在的食品、药品和危险品等货物的分组包装问题属于带冲突关系的装箱问题( BPPC),其优化目标是在满足货物间冲突限制的前提下完成装箱操作,并最小化使用货箱的数量。本文从实际需求出发,基于货物之间的冲突关系、装箱顺序和货箱容量等约束建立相应的数学规划模型;随后设计了求解BPPC问题的启发式算法,算法通过迭代求解最大团结构实现货物间冲突关系的消去,根据当前货物最大团采用改进降序首次适应算法( FFD)完成货物装箱操作,并通过“洗牌”策略对已有装箱方案进行局部优化;最后,针对Iori算例数据,将以上算法与基于图着色的启发式算法进行比较分析,结果表明,本文算法是求解BPPC问题更为有效的方法。
現實物流活動中大量存在的食品、藥品和危險品等貨物的分組包裝問題屬于帶遲突關繫的裝箱問題( BPPC),其優化目標是在滿足貨物間遲突限製的前提下完成裝箱操作,併最小化使用貨箱的數量。本文從實際需求齣髮,基于貨物之間的遲突關繫、裝箱順序和貨箱容量等約束建立相應的數學規劃模型;隨後設計瞭求解BPPC問題的啟髮式算法,算法通過迭代求解最大糰結構實現貨物間遲突關繫的消去,根據噹前貨物最大糰採用改進降序首次適應算法( FFD)完成貨物裝箱操作,併通過“洗牌”策略對已有裝箱方案進行跼部優化;最後,針對Iori算例數據,將以上算法與基于圖著色的啟髮式算法進行比較分析,結果錶明,本文算法是求解BPPC問題更為有效的方法。
현실물류활동중대량존재적식품、약품화위험품등화물적분조포장문제속우대충돌관계적장상문제( BPPC),기우화목표시재만족화물간충돌한제적전제하완성장상조작,병최소화사용화상적수량。본문종실제수구출발,기우화물지간적충돌관계、장상순서화화상용량등약속건립상응적수학규화모형;수후설계료구해BPPC문제적계발식산법,산법통과질대구해최대단결구실현화물간충돌관계적소거,근거당전화물최대단채용개진강서수차괄응산법( FFD)완성화물장상조작,병통과“세패”책략대이유장상방안진행국부우화;최후,침대Iori산례수거,장이상산법여기우도착색적계발식산법진행비교분석,결과표명,본문산법시구해BPPC문제경위유효적방법。
In real distributions , there are a great many of packing problems with food , medicine and hazardous material named bin packing problem with conflicts , whose objective is to minimize the number of used bins and has to satisfy the conflict constraints among the items .To solve the problems , the mathematical model of BPPC is proposed based on the conflict relationship among the items , packing order and capacity constraints of the bins;and then a heuristic algorithm is designed for solving BPPC , the algorithm computes the maximum clique struc-ture iterately to eliminate the conflict relationship among the items , runs the packing operation according to the items corresponding to the current maximum clique structure , and a local search methed named shuffling strategy is applied to further optimize the current packing results;lastly the simulation experiment is carried out on Iori ’ s strandard instances compared with the heuristic algorithm based on the graph coloring model , the computational results demonstrate that the heuristic algorithm in this paper is more feasible and effective .