透過您的圖書館登入
IP:3.22.248.208
  • 會議論文

産品年度保固料件需求預測模式之研究

The Demand Forecasting Model of Yearly Products Warranty Materials

摘要


産品年度保固是企業的重要工作之一,而産品年度保固料件之需求預測,涉及産品年度保固之成效之良窳。由於目前實務之作法,係以過去的經驗值來估算,常常無法正確估計産品年度保固料件之需求,因此需要一套有效的産品年度保固料件之需求預測方式。因此本研究乃以結合針對少量資料的灰色理論以及具效率的指數平滑法,建構産品年度保固料件預測模式,並以某維修工廠産品年度保固料件為例,分析其預測效果。本研究之預測結果與其原始預測資料,比較發現其平均絕對誤差(MAPE)、絕對誤差檢驗均有良好之改善。 最後依據本研究所提出結合灰色理論、指數平滑法及趨勢指數平滑法91種演算法之産品年度保固料件年度需求預測模式,建置一套以Delphi5應用程式為基礎,所寫之預測軟體,供承辦人員藉由歷史資料的輸入,得到未來可能的需求預測值,可以即時得到資料,並加以運算,使預測之效率及效果提高。

並列摘要


Yearly products warranty is the most important job of businesses; The demand forecasting results of yearly products warranty materials affect yearly products warranty. The forecasting way we do presently is based on the past experiential values, which cannot estimate yearly products warranty materials correctly. Therefore; we need a set of the demand forecasting model. This research combines grey theory, exponential smoothing and trend exponential smoothing to construct demand forecasting model. This research analyzes some company's the demand forecasting results of yearly products warranty materials. Compared with The results of this research, the original forecasting data have the better MAPE and absolute error index. This research combines 91 kinds of algorithms of grey theory, exponential smoothing and trend exponential smoothing to build a set of forecast software based on Delphi5. This forecast software can let undertake personnel inputs historical data to get demand forecasting values, which can increase forecasting efficiency.

延伸閱讀