基于区间估计方法的零星需求电力物资协议采购方案

Prediction-interval-based Purchase Strategy in Agreement for Slow-moving Electric Power Materials

针对零星需求电力物资的库存管理问题,设计了一个基于区间估计方法的协议采购方案。首先提出一个对历史无需求产品构成的产品池建立未来总需求率区间估计的方法;然后在不同的参数条件下对这种方法进行可靠性检验,确定区间估计方法适用的参数范围;最后选取符合参数条件的电力物资,针对其中特定期间内零需求的物资,采用此方法建立未来总需求率的区间估计,并结合其价格水平,构建协议采购方案。此方案提高了采购效率,实现了零库存成本、高服务水平的库存管理目标。

The dilemma in inventory management of electric power materials with slow-moving demand is defined and resolved by purchase in agreement. A proposed prediction interval of estimated aggregate future demand rate is constructed for a product pool in which all products have no demand historically. A simulation study examines the reliability of the methodology across various parameters. The adaptable pa- rameters are employed in the selection of electric power materials. The interval of demand rate for those with an observed demand of zero in given period out of all selected materials is estimated, The prediction interval of aggregate demand rate and the real price level are combined to calculate the sum for funding the demand of those materials in a time unit. The purchase contract price in agreement is decided accordingly, by which electric power enterprises are allowed to boost purchase efficiency and achieve inventory management objectives of both zero cost and high service level.