Available-to-promise modeling for parts arriving uncertainty with fuzzy goal programming

[[abstract]]本論文主要之目的在探討當原物料到貨量不穩定下，製造商之可允諾量 最佳排程的模式規劃。傳統的可允諾量是根據主排程(Master Production Schedule, MPS)的結果Chen , Zhao & Ball, ( 2001)，由前一期的期末存貨水準加 上當期預計生產完成的產品加總而得的，換言之，傳統的可允諾量只是單純 地計算製造商尚未被消耗完的可供給額度，而業務部門便根據可允諾量數量 來判定新訂單的接受與否。 然而，在真實的生產環境中，會因為各種不確定因素的相互影響，以致 整個決策與供應系統存在著不確定性，而在決策者面臨不確定的狀態下，其 思考的方向及喜好程度也難以說明。因此利用模糊集合理論(Fuzzy Sets Theory)，將真實狀況所產生的不確定性融入數學模式中，來幫助決策者做判 斷是一項較有說服力的方式。 本研究是運用ATP(available to promise)的弁遄A它能提供所能供應之產能 與物料等製造資源之資訊，回覆顧客訂單的需求，在接單後組裝的生產模式 (ATO, Assemble-to-order)下，則根據可獲得的物料和生產產能作為回覆顧客訂 單的依據，在此概念下，以作業研究的整數規劃方法建構模式，在模式中考 量關鍵物料可能有缺貨之虞，因此以模糊數學的計算方式來定義出可接受的 到貨水準，結合目標函數利潤最大化尋求其最佳排程。 研究中使用兩種不同之求解步驟分別為傳統利潤極大化與模糊目標規劃 方式，探討在不同的到貨水準下應使用何種方式進行可允諾量排程，比較出 其不同之優缺點。最後，為了妥善使用期初能，加入考量未用產能成本的觀 念，讓企業本身在考量到貨不穩下，也能完成訂單，並有效地使用產能。[[abstract]]In this study, we analyze original supplies parts arriving uncertainty, manufacturers how to planned ATP(available-to- promise) quantity , A traditional ATP is according to MPS (Master Production Schedule) ,it is expect by previous final inventory level as issue products finished to produce add always and take to add. in other words, traditional ATP just calculate how much can use by not yet used parts, and the business segments judge the acceptance of the new order or not according to the quantity of promising amount. But in the true production environment, various of uncertain factors will influence each other, so that the whole supply chain and decision system with have uncertainty, under the circumstances the decision-maker faces the uncertain state, we difficult to explain its direction of thinks. So use the fuzzy set theory incorporate the true uncertainty produced state in the mathematics model , it will help decision-maker judge decision. In this study we use two different solution step, one is maximization profit scheduling solution another is fuzzy goal programming solution, to confer which one weed to use in different arriving level, to know which different between each other. Finally, in order to can use properly initially capacity we add Virtual profit to the model, and to calculate solution