系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2015年
7期
1678~1688
,共null页
能力扩张 随机 scenario 粒子群优化 双目标规划
能力擴張 隨機 scenario 粒子群優化 雙目標規劃
능력확장 수궤 scenario 입자군우화 쌍목표규화
capacity expansion;stochastic;scenario;particle swarm optimization;bi-objective programming;
能力扩张问题是指在不同计划期, 根据不断增长的市场需求, 调整生产能力, 使产能与需求合适匹配以寻求企业盈利的最佳表现. 能力扩张涉及长期的产能投资, 是生产领域的战略决策问题. 当面对不确定需求时, 这一决策任务变得十分复杂, 需要从回报与风险两个角度评估能力扩张方案的性能. 本文基于Scenario树描述多阶段的随机需求, 从而建立随机环境下能力扩张问题的Scenario决策模型. 其中考虑两个决策目标: 最大化利润和最小化期望下方风险, 因而形成了一个双目标规划问题. 为求解该问题, 基于二进制粒子群优化技术, 提出了双目标优化的粒子群算法. 算例表明该算法可以得到近似Pareto前沿, 且能揭示利润与风险的同向变化关系.
能力擴張問題是指在不同計劃期, 根據不斷增長的市場需求, 調整生產能力, 使產能與需求閤適匹配以尋求企業盈利的最佳錶現. 能力擴張涉及長期的產能投資, 是生產領域的戰略決策問題. 噹麵對不確定需求時, 這一決策任務變得十分複雜, 需要從迴報與風險兩箇角度評估能力擴張方案的性能. 本文基于Scenario樹描述多階段的隨機需求, 從而建立隨機環境下能力擴張問題的Scenario決策模型. 其中攷慮兩箇決策目標: 最大化利潤和最小化期望下方風險, 因而形成瞭一箇雙目標規劃問題. 為求解該問題, 基于二進製粒子群優化技術, 提齣瞭雙目標優化的粒子群算法. 算例錶明該算法可以得到近似Pareto前沿, 且能揭示利潤與風險的同嚮變化關繫.
능력확장문제시지재불동계화기, 근거불단증장적시장수구, 조정생산능력, 사산능여수구합괄필배이심구기업영리적최가표현. 능력확장섭급장기적산능투자, 시생산영역적전략결책문제. 당면대불학정수구시, 저일결책임무변득십분복잡, 수요종회보여풍험량개각도평고능력확장방안적성능. 본문기우Scenario수묘술다계단적수궤수구, 종이건립수궤배경하능력확장문제적Scenario결책모형. 기중고필량개결책목표: 최대화리윤화최소화기망하방풍험, 인이형성료일개쌍목표규화문제. 위구해해문제, 기우이진제입자군우화기술, 제출료쌍목표우화적입자군산법. 산례표명해산법가이득도근사Pareto전연, 차능게시리윤여풍험적동향변화관계.
Capacity expansion is the problem that an enterprise adjusts its production capacity to meet the continuously increasing market demand, and appropriately matches the production capacity and the demand to pursue the optimal performance of the profit. Capacity expansion concerns the investment of the production capacity during a long term, such that it is a strategic problem in production. When confronted with uncertain demand, the decision becomes very complicated. The decision maker need evaluate the performance of the capacity expansion scheme simultaneously considering return and risk. This paper describes the stochastic demand information based on scenario tree and formulates scenario-based programming model of capacity expansion under stochastic environments. The objective of the model is two-fold: maximizing the expected profit and minimizing the expected downside risk, which forms a bi-objective programming problem. To solve the model, a particle swarm optimizer of bi-objective optimization is proposed, which is developed based on the binary particle swarm optimization. The numerical study shows that the proposed algorithm can obtain an approximate Pareto frontier, as well as reveal the positive correlation between the profit and the decision risk.