在需求預測的規劃與應用上，模糊時間序列模式從連績兩期資料間的模糊關係推估下一期之預測值。雖然模糊時間序列提供了一個不概念的預測工，然而對於季節性資料的預測皆析，模糊時間序列的分析步驟尚缺乏一套穩健且精準流程。因此本研究從模糊時間序列模式著手，結合統計方法之概念轉換季節性資料成為較稱定之資料，並適當修正模糊時間序列模式，使得預測的應用績效更精良。最後，將舉一實例說明之。

In the demand forecasting, the fuzzy time series model has its own advantages to be a forecasting tool, it is still lack of a completely and robustly analysis procedure for seasonal time series. In order to cope with such problem, we combined the statistical methods into fuzzy time series model to transform the seasonal data to be a stationary type to improve extrapolative performance in the fuzzy time series model. Finally, an example is provided for illustration.