TFT-LCD產業競爭激烈，TFT工廠也隨著玻璃世代增加而日趨大型化，能源需求量也隨之倍數增長，TFT-LCD產業也已成為台灣高耗能產業之一。在全球能源需求量日益增加、能源費用高漲及國際間節能減碳意識興起之下，TFT-LCD企業主管也基於綠色管理理念，開始進行產品耗能指標管理。產品耗能指標及主要耗能設備效率指標可作為推動節能目標之參考依據。 高階管理者可在若在資源有限的條件下，建立能源管理系統及綠色製程供應鏈，以有效地改善能源使用模式，並真正地落實能源查核與管理，再結合產品品質管理與綠色產品研發，以提升企業之競爭力及獲利力。 TFT-LCD單位產品耗能指標之變數很多，製程設備的耗電量佔約全廠區用電量的三分之一，且製程設備耗電量較不受外界環境之影響，因此，本研究使用TFT-LCD廠產能投入數據及製程設備歷史運轉用電量為研究條件，再利用時間序列分析法為預測方法，建立TFT-LCD製程耗能模型，製程設備單位耗能指標可作為單位產品耗能指標之先行研究，並做為用電量的預測及工廠管理指標之一，以期落實TFT-LCD產業之綠色生產、提升企業競爭力及企業永續經營發展。 經由本論文研究可以證實採用時間序列分析法對於建立TFT-LCD製程耗能指標模型有很好的配適度並且各項係數都有極佳的顯著性，且在進行樣本外預測製程耗能時皆有很好的精準度，整體而言，ARMA線性預測模型可以精確地掌握TFT-LCD製程耗能指標模型。 本論文研究證實時間序列分析法可以建立TFT-LCD製程耗能指標模型，並且有很好的配適度及預測力，未來還可再加入其他外在環境變數，做更精確的預測模型分析，再建立TFT-LCD單位產品耗能指標或TFT-LCD單位產品耗水指標。;The TFT-LCD industry is highly competitive. To have scale economy, the TFT glass factories have increased its production size. This has led the energy demand growing exponentially. The TFT-LCD industry has become one of the high energy-consuming industries in Taiwan. Due to the increasing global energy demand, the rising energy costs and international awareness of energy saving, the executives of TFT-LCD industries have paid attention to the functions of green management and green products. With the conditions of limited resources, the establishment of energy management systems and green supply chain process can effectively improve the patterns of energy use. The implementation of energy management and green product development would enhance the competitiveness and profitability of enterprises. The indicators of energy consumption of TFT-LCD products have many factors. The product processing accounts for more than one-third of the total energy consumption, which is not affected by the external environment like weather temperature. The purpose of this thesis uses the data of electricity consumption to find the relationship between the processing of energy consumption and the output of TFT-LCD products. By using the time series analysis approach, our empirical results indicate that ARMA model has a good prediction power to construct the relationship. Our study confirms that the time series analysis can be used to build up the energy consumption models for the manufacturing process of TFT-LCD products. Furthermore, the approach can be extended to find the relationship between the water consumption and the manufacturing process in the future work.
