模糊時間數列預測模型普遍存在兩項缺點：(1)單一預測值，非區間侓，無法提供決策者更多相關資訊；(2)當突發狀況發生時，歷史資料無法即時反應，造成預測誤差。針對上述問題，本文發展一綜合型非時變模糊時間數列預測模式，結合模糊時間數列預測與專家判斷，以模糊預測值(模糊數)提供決策者相關訊息。本文先以一範例測試，並與其他模糊預測模式比較，結果顯示；本模型預測結果比其他預測模式較佳，再以楠梓加工區被動元件製造公司產品需求預測實例，說明植模式之實用性。

Various fuzzy time-series forecasting models have been proposed. However, most of them have the following problems: (1) the forecasting values are a single value; (2) When unexpected conditions happen, the historical data can not be used to predict the situations properly. In order to solve the problems mentioned above, this paper proposes an integrated time-invariant fuzzy time series forecasting model that combines fuzzy time series forecasting method and experts’ judgment to predict the time series. The fuzzy forecasting values (fuzzy number) provide decision makers useful information. One example is used to validate the accuracy of the proposed model. The results indicate the proposed model is better than the other existing models. At last, the proposed model is implemented in a real-word case.