零工型工廠交期設定模式之構建

Abstract

提高訂單交期績效 (Due-Date Performance) 及縮短製造前置時間 (Manufacturing Lead Times)，可說是製造業最重視之生產績效因子。因 而在客戶詢、報價及業者產銷協調規劃過程中，若能準確、合理的設定訂 單發放日期及交貨日期，將能使工廠準時完工交貨及降低在製品 (WIP) 存貨的能力大增。基於此一理想，本文綜合過去有關零工型工廠 (Job Shop) 交期設定方法研究上之缺失：(1) 訂單發放日期已知，著重於訂單 之流程時間估算 (Flowtime Estimation)；(2) 訂單交貨日期已知，著重 於訂單發放控制 (Order Release Control)。首先就上述缺點來作改善並 構建發展新的整合式交期設定方法 (Integrated Due-Date Assignment Method) ，此整合式交期設定法同時融合了訂單發放控制及流程時間估算 二項功能。模擬驗證結果說明了整合式交期設定法不僅提高了訂單等候時 間估算之準確度，而且也提昇了交期績效。茲因預測、估算的方法所使用 的加工資訊大都為加工歷史資料 (Historical Data) 及一些控制參數 (Control Parameter)，因而即使比較優適的整合式交期設定方法也會造 成部份訂單之流程時間被高估或低估 (Over/Under Estimated) 之情形。 為了解決此問題，本研究乃構建另二類訂單交期設定模式，此二種設定模 式分別為：(1) 以產能限制資源 (Capacity Constraint Resource, CCR) 為基磐之交期設定法，此交期設定法的精神乃延伸於限制理論 (Theory of Constraint, TOC) & DBR (DRUM-BUFFER-ROPE) 排程法之理念。本研 究同時以此交期設定法與 DBR 排程法 (DISASTER) 作比較，驗證結果證 明本方法較為優適，而且也改善了 DBR 應用上之缺失；(2) 混合排程交 期設定法 (Hybrid Scheduling Due-Date Assignment Method)，此交期 設定法的精神則延伸於前推有限產能排程 (Forward Finite Capacity Scheduling) 及後推有限產能排程 (Backward Finite Capacity Scheduling) 之功能。希望藉由此三種不同交期設定方法之應用，讓製造 業者在訂單規劃時，能依公司不同之產銷型態，選擇不同之交期設定模式 ，使業者能夠準確、快速、合理的設定訂單交貨日期及發放日期。 In the past, many production scheduling methods were developed to achieve high due-date and short manufacturing lead times performance. Although having attained valuable insight into due- date assignment problems involving the job shop manufacturing system, previous researches have neglected the characteristics that are essential for effective due-date performance: order release function focus and thinking on the integrating the order release control methods with due-date assignment rules. Therefore, in this study, we propose an Integrated Due-Date Assignment Method to overcome the above characteristics. This study integrates order release control methods with due-date assignment rules and assesses its impact to the accuracy of interoperation time estimation and performance of due-date assignment. Simulation results in this study indicate that integrating the order release control methods with due-date assignment rules will significantly enhance not only the accuracy of interoperation time estimation, but also the performance of due-date assignment rules. Since the interoperation time is estimated using some historical data and control parameters, some of the values will be under/over estimates, which will lead to some orders being delivered late/ early unless appropriate remedial action is taken. Therefore, in this study, we also propose the two due-date assignment methods to overcome the flowtime under/over estimation problems. The two due-date assignment methods are: (1) Due-Date Assignment Method Based on CCR (Capacity Constraint Resource, CCR), which is derived from TOC (Theory of Constraint) and DBR (DRUM-BUFFER- ROPE). Also, an illustrative example demonstrates the feasibility of the proposed method and the results are compared with the DBR method; (2) Hybrid Scheduling Due-Date Assignment Method, which is derived from the function of forward finite capacity scheduling and backward finite capacity scheduling. Finally, we hope the proposed methods can help manufacturers to achieve high due-date performance.