有限需求信息下基于最大熵原理的风险厌恶库存模型

发布时间：2018-05-08 12:04

  本文选题：库存模型 + 最大熵原理　； 参考：《东北大学学报(自然科学版)》2016年10期

【摘要】：针对风险厌恶的库存决策者,建立了基于条件风险值(CVaR)的单周期库存模型.在仅知需求区间、均值和方差信息情况下,采用最大熵原理估计了两种条件下的需求分布.结果显示,在仅知需求区间、均值和方差信息时,决策者应分别采用均匀和指数分布作为潜在的需求分布.在此基础上,进一步推导了基于CVaR的库存订货策略及其绩效情况.模拟结果表明,同真实需求分布下的最优情况相比,基于最大熵原理的库存策略虽然会导致绩效损失,但损失比例很小,表明基于最大熵原理的订货策略具有良好的鲁棒性.
[Abstract]:A single cycle inventory model based on conditional risk value (CVaR) is established for risk-averse inventory decision makers. The maximum entropy principle is used to estimate the distribution of demand under the condition of only knowing the information of demand interval, mean value and variance. The results show that when the information of demand interval, mean value and variance are only known, the decision makers should adopt the uniform and exponential distribution as the potential demand distribution respectively. On this basis, the inventory ordering strategy based on CVaR and its performance are further deduced. The simulation results show that the inventory strategy based on the maximum entropy principle can lead to the loss of performance, but the loss ratio is very small, which indicates that the order strategy based on the maximum entropy principle has good robustness.
【作者单位】： 东北大学工商管理学院;
【基金】：国家自然科学基金资助项目(71372186) 教育部人文社会科学研究一般项目(11YJC630165) 中央高校基本科研业务费专项资金资助项目(N150604005)
【分类号】：F274
，

本文编号：1861328