某型导弹引信备件需求预测模型研究  

Research on Spare Parts Demand Forecasting Model of a Missile Fuze

作　　者：任喜[1] 赵建军[1] 孙靖杰[1] 杨利斌[1] 

机构地区：[1]海军航空工程学院,山东烟台264001

出　　处：《计算机仿真》2014年第10期67-71,共5页Computer Simulation

基　　金：国家自然科学基金资助项目(60478053)

摘　　要：准确的备件需求预测在备件库存管理中尤为重要,但是备件间歇性需求的特点以及需求的历史数据不足使得备件需求预测特别困难,在一些部门,导弹备件库存管理水平决定了设备如何使用和维修。为解决上述问题,建立一种新的预测方法,综合评价备件需求预测的自相关与解释变量的关系:通过回归分析确定由解释变量引起的非零需求,判别一个非零需求发生的原因,进而得到修正后的非零需求序列的预测结果,最后通过LTD估计,得到具体的非零需求预测值。通过理论分析以及实例结果均证明,改进方法在预测精度上优于其它方法,为导弹备件预测研究提供一定的参考与借鉴。Accurate demand forecasting is particularly important in the spare parts inventory management, but the characteristics of the intermittent demand and the lack of historical data make spare parts demand forecasting particu- larly difficult, in some sectors, missile spare parts inventory levels also determine on how to use and maintenance the equipment. Starting from the above three difficulties, a new prediction method is established for comprehensive eval- uation of the relationship between the demand forecast from spare parts associated with the explanatory variables ： De- termine the non - zero demand caused by the explanatory variables in the regression analysis, discriminate the cause of a nonzero demand, and then get the corrected forecasting result of the sequence of nonzero demand, get specific nonzero demand forecasting by the final LTD estimate. Through theoretical and example analysis, the result is superi- or to other methods in the prediction accuracy, and this new method can provide a reference for the prediction of mis- sile parts.

关 键 词：备件预测 精度预测 指数平滑法 综合预测法 

分 类 号：TJ760.7[兵器科学与技术—武器系统与运用工程]