具有遗憾值约束的鲁棒供应链网络设计模型研究

考虑不确定性环境,研究战略层次的供应链网络鲁棒设计问题,目标是设计参数发生摄动时,供应链性能能够保持稳健性。基于鲁棒解的定义,建立从上游供应商选择到下游设施选址-需求分配的供应链网络设计鲁棒优化模型;提出确定遗憾值限定系数上限和下限的方法,允许决策者调节鲁棒水平,选择多种供应链网络结构;通过模型分解与协调,设计了供应链节点配置的禁忌搜索算法。算例的计算结果表明了禁忌搜索算法具有良好的收敛特性,以及在处理大规模问题上的优越性;同时也反映了利用鲁棒优化模型进行供应链网络设计,可以有效规避投资风险。

Research on Supply Chain Network Design Model with Regret Value

Based on estimated potential customer demand,a company selects its suppliers from qualified candidates,determines locations,capacities in its manufacturing plants and distribution centers,and decides the transportations among them and to customers.A successful supply chain network design,which enables company to integrate its suppliers,plants,distribution centers,and the customer zones efficiently,provides sufficient flexibility in future operations in the dynamic markets,and keeps construction,purchasing,transportation,and operations costs at a reasonable low level,eventually reinforces a company’s long term competitiveness. The current literature independently examines different problems,such as supplier selection,facility location and capacity designs,based on the deterministic assumption;that is,all parameters,such as marketing demand,investment in facility,distance between facilities,costs and prices,etc.,are assumed constant.A linear programming or heuristic algorithms have been used to reach a optimal solution under the deterministic assumption.However,this assumption may not be true in real life and design parameters may change over time.In practice there may exist many possible scenarios where design parameters(such as demands and costs)may take different values.The optimal solution based on the deterministic assumption may not be good for all scenarios.Therefore,it is desirable to consider parameter uncertainty in design and find a solution that can perform well under any possible scenarios This paper investigates the strategic robust supply chain network design problem of keeping stable performance when design parameters fluctuate.In the first part,based on the definition of robust solution,a robust optimization model is built for integrated supply chain network design that consists of supplier selection problem and facility location-distribution problem.In the second part,the method is used to determine the upper bound and the lower bound of the restriction coefficient on regret values.The decision-maker is allowed to adjust the robustness level and has options to design supply chain networks.In the third part,by the decomposition of the optimization model,tabu search algorithm is proposed for supply chain node configuration. At last,the contrastive test result validates that the performance of tabu search algorithm is outstanding on convergence and computational time.Meanwhile,it is indicated that the robust optimization model can reduce investment risks effectively when it is applied to supply chain network design.