因多晶矽太陽能電池片製造廠產出之轉換效率分佈是屬非均一性且具隨機性，即使參考歷史產出資料作為轉換效率分佈的假設基礎下，實際上產出的轉換效率分布結果也常出現相左之情況。故，衍生出允許在訂單上作調配之補救措施：可替代性，所謂的替代性指的就是電池片之轉換效率具有可向下替代之特質。這也讓生產規劃者須同時面臨產品組合選擇、投產產品數量、投產時程及訂單配置這四項因素所交織的問題。然而，為了應對市場價格波動及跌價所造成的存貨壓力，生產規劃者必須找出適當的投產組合、數量及時程，在滿足訂單的條件下，控制庫存水準。故本研究將藉由規劃求解之工具來達成此目的並改善解決作計畫之繁雜程序。 Excel規劃求解的增益集讓多重期間生產規劃及產品組合選擇等這方面的線性規劃問題在求解上更具有效率。故，本研究將因應Excel的資源限制，建構一個線性規劃之數學模式，在滿足訂單的條件下，並探討訂單可替代性的狀況，找出合適的產品組合、投產數量及時程，並求出最小總投產量以控制庫存水準。最後，比較現行多晶矽太陽能電池片製造廠之實際做法，以驗證本數學模式之實用性與可行性。 ;Due to the multi-crystalline silicon solar cell products manufacture the efficiency distribution is not only the non-homogeneity but also a randomness. Based on the situation, in order to meet the order request and reduce the inventory level, the order configuration of the efficiency can be upgraded and replaced from the low to high. This paper studies a linear programming (LP) problem of solar cell products selection and multi period scheduling which based on the efficiency distribution is randomness and order configuration can be upgraded and replaced. With this mathematical methodology to make a best production plan decision based on different considerations including minimum cost, minimum the inventory level and fulfillment of customer efficiency request, quantity and delivery date.
