基于信息更新的多产品两阶段订货风险决策模型  

Product-portfolio ordering analysis with update information in the two-echelon: a risk decision-making model

作　　者：周艳菊[1] 邱菀华[2] 王宗润[1] 

机构地区：[1]中南大学商学院,长沙410083 [2]北京航空航天大学经济管理学院,北京100083

出　　处：《系统工程理论与实践》2008年第1期9-16,共8页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(70771114);高校博士点专项科研基金(20030006009)

摘　　要：借鉴金融工程领域广泛应用的条件风险值法,以及基于布朗运动的贝叶斯预测方法,建立两阶段多产品订货风险决策模型,用数值分析对模型进行了检验,发现它基本反映了真实的决策过程和决策者心理.由于所建的模型最终可表示为线性规划问题,因此即使是大规模的产品组合问题也可以借助工具软件求解.The paper proposed a two-echelon optimal ordering model for multi-products with Conditional Value-at- Risk（CVaR） which is used popularly in the field of financial engineering. Bayesian forecasting under Brownian motion is adopted as the method of updating information, for it more adapts to the strong stochastic character of single-period product. The model is then tested by simulative data, the outcome of which follows the real decision-making behavior completely. Moreover the model can be formulated as a linear programming problem which is solved by the technique software easily.

关 键 词：多产品订货 两阶段 条件风险值 布朗运动 贝叶斯预测 

分 类 号：F273.17[经济管理—企业管理]