基于风险测度的供应链鲁棒建模与策略研究

Uncertainty is an inner phenomenon in the supply chain operation, which affects the supply chain performance in a serious way. Empirical investigations show that the decisions adopted may not always coincide with that made based on expected profit or cost because of the different preferences for the supply chain performance risk arises from uncertainty. Some theoretical approaches such as mathematical programming, robust optimization, risk measurements and so on in this project will be used to develop some supply chain robust operation models which take supply chain performance indicators measured with different risk measures as objective functions, various kinds of operational settings as constraint conditions, and many operations as decision variables. First, some robust models of multi-echelon supply chain will be establised based on single stage supply chain robust model of single period and multi-period; Second, a supply chain robust optimization model based on product sales life cycle will be developed with the consideration of different risk preferences of decision-maker in various phase of the cycle; At last, the supply chain robust optimization models based on value performance will be built by measuring supply chain operation from the view point of financial. According to the above models, supply chain robust decisions under uncertain environment will be proposed, and the impacts of risk preference factors on supply chain operations will be analyzed. The research results of this project will provide some supports for the decision-maker with different risk preferences to against uncertain disturbances in operation.

不确定性是供应链运作中的内在现象，严重影响供应链的性能。实证研究表明，由于决策者对不确定性导致的绩效风险的偏好态度不同，执行的决策并不总是与基于期望利润/成本的决策一致。本项目将在分析不确定环境、风险决策和鲁棒建模准则基础上，运用数学规划、鲁棒优化、风险测度等理论方法，建立以不同风险测度衡量的供应链绩效为目标函数，各种运作环境为约束条件，若干运作为决策变量的供应链鲁棒运作模型。首先，在单周期和多周期单级供应链鲁棒建模基础上，建立基于风险测度的多级供应链鲁棒优化模型；其次，考虑不同销售周期阶段决策者的风险偏好可能不同，建立基于产品销售生命周期的供应链鲁棒优化模型；最后，从财务角度衡量供应链运作情况，建立基于价值绩效的供应链鲁棒优化模型。针对上述模型，求解不确定环境下供应链鲁棒策略，分析风险偏好因素对供应链鲁棒运作的影响。本项目研究成果将对具有不同风险偏好的决策者应对不确定性扰动提供决策支持。

不确定性是供应链运作中的内在现象，严重影响供应链的性能。实证研究表明，由于决策者对不确定性导致的绩效风险的偏好态度不同，执行的决策并不总是与基于期望利润/成本的决策一致。在此背景下，本项目通过分析不确定环境、风险决策和鲁棒建模准则，运用数学规划、鲁棒优化、风险测度等理论方法，建立以不同风险测度衡量的供应链绩效为目标函数，各种运作环境为约束条件，若干运作为决策变量的供应链鲁棒运作模型。首先，对供应链风险决策准则和鲁棒运作机理进行了理论研究，在此基础上，针对单级供应链系统，分别研究了风险厌恶下的单周期和多周库存鲁棒运作问题；其次，将研究扩展到由制造商、批发商、零售商等构成的多级供应链系统，考虑供应链运作中的不确定性及由此导致的绩效风险问题，研究了基于风险偏好的多级供应链鲁棒运作问题；再次，从财务绩效角度考虑供应链运作中的不确定性风险和资本支出问题，采用风险驱动树框架分析企业价值创造过程中的若干运作环节和影响因素，建立了基于价值绩效的供应链鲁棒优化模型；最后，对上述模型进行仿真实验，给出了相应的鲁棒运作策略，验证了所建模型及鲁棒策略的有效性，并分析了不同风险偏好参数对供应链运作绩效的影响。研究表明，供应链运作中不确定性问题的存在虽然会导致相应的绩效损失，但损失值很小，表明基于本项目研究所得到的鲁棒策略能够有效抑制供应链系统中的不确定性扰动。本项目研究成果一方面丰富了供应链鲁棒性问题研究理论，另一方面将对具有不同风险偏好的决策者应对不确定性扰动提供有效决策支持。