本研究以TFT-LCD(thin film transistor–liquid crystal display)產業鏈面板製造商為例,探討多廠區產能擴充決策與產能分配規劃之問題。由於該產業具有許多生產特性與限制,例如複雜的產品階層結構造成了產品種類繁多且應用廣,以及TFT-LCD產業有多階層、多世代與多廠區共存的生產鏈結構,再加上快速成長和變動的產品需求,使得產能規劃在TFT-LCD產業變的格外重要。 本研究針對未來需求具有不確定性之環境,提出兩階段隨機產能規劃模式(Two-stage Stochastic Capacity Planning Model),並考量購買瓶頸機台與附屬工具(光罩)之產能擴充決策,模式目標為最大化淨現值 (Net present value; NPV)。本研究以TFT-LCD面板產業為例,藉由Monte Carlo Simulation實驗以模擬三種變異情境下的未來需求,進而比較需求不確定性與需求確定性之產能規劃模式,從實驗結果可知當需求波動劇烈時,需求不確定性規劃模式的求解結果獲利較高,因此規劃的產能擴充與產能分配決策會較佳。最後再利用不同參數水準進行敏感度分析,以了解不同參數水準變化對產能規劃結果及淨利潤等財務指標的影響。
This study explores multi-site capacity expansion decision and capacity allocations through example of TFT-LCD industry. This study has proposed two stage stochastic capacity planning model under uncertainty of demand, and gives the capacity expansion decisions based on purchasing bottleneck machines and auxiliary tools (photo mask) to reach the model target of maximizing NPV (Net present value). This study simulated the future demand under 3 different conditions via Monte Carlo experiment, and further compared the capacity planning model between demand certainty and demand uncertainty. Ultimately, sensitivity analysis is also carried out using different parameters.