| Summary: | 碩士 === 國防管理學院 === 資源管理研究所 === 95 === The demand forecasting of yearly military weapon repair materials affects the availability of military weapon equipments. The present forecasting method bases on the experiential judgement, which cannot forecast yearly military weapon repair materials correctly. Consequently, we need a good demand forecasting model for yearly military weapon repair materials.
This study combines grey theory, exponential smoothing and trend exponential smoothing to construct the demand forecasting model. To analysis the feasibility of the proposed method, the yearly demand of 997 military weapon repair materials items with the original forecasting results from 2001 to 2006 are collected and the the repair materials are classified by ABC method.
The forecasting results of the proposed method show better performance than the original forecasting results in performance indices including total evaluation indications, total error cost and mean absolute error detecting. Among the ABC classifications, classification A shows the best performance indices.
Finally, this study use Delphi 5.0 to build a forecasting software based on grey theory, exponential smoothing and trend exponential smoothing. The users of this software can input data interactively or in batch to get the demand forecasting data efficiently.
|
|---|
