不同协作模式下的多级供应链动态优化

Dynamic Optimization of Three-Stage Supply Chain under Different Collaborative Modes

在一阶自回归需求和定购点策略的假设组合下研究了多级供应链的协作模式,即通过对成员企业定购策略、提前期、需求模式等信息的共享来优化供应链整体绩效。与以往研究不同,文中引入的成本目标函数涉及相互冲突的定购量和实际库存动态变化,并针对全局优化模式和无私零售商模式下的多级供应链绩效分别进行了详细分析。与全局优化模式相比,无私零售商模式对信息共享的深度放宽且对总利润再分配机制的要求简化。进一步的数量研究表明,无私零售商模式虽非最优,却可在制造商不必作出定购策略再设计的情况下,得到比传统定购模式更合理的整体绩效。

The dynamic optimization problem of a three-stage supply chain composed of one supplier,one manufacturer,and one retailer is discussed.In such a supply chain,it is assumed that market demands follow first-order auto-regressive process and the members share relevant information,e.g.,order policies,lead times,and demand models.Order models for different members are discussed.Different from the existing studies,new objective is used,in which inverse changes between the bullwhip effect and magnification ration of actual stock are considered.Then,system performance is analyzed under two collaborative modes：（1） global optimization mode,where the manufacturer and retailer follow smoothing order point policy and global information sharing is required;and（2） altruistic retailer mode,where the manufacturer follows general order point policy and the retailer adjusts its proportional controller for achieving global system optimization without selfishness.It is shown that,under altruistic retailer mode,although quasi-optimal supply chain performance is achieved,information sharing and gross profit re-allocation are more operable than those under global optimization mode.Compared with that obtained when general order policy are adopted,more reasonable system performance can be obtained under altruistic retailer mode without redesigning order policy of the manufacturer.