反向模擬派工法則

現行的製造產業已轉變成快速生產、產量少而樣式多，在如此的生產環境下如何降低總延遲時間並提昇達交率成為主要的課題，同時也是企業競爭力的所在。因此有許多相關研究討論排程的派工法則，針對不同的情形提出不同的解決方法。本研究主要是探討Job Shop的排程問題，同時考量交期與容許等待時間這兩個因素進行排程。為達成此目標，我們提出以「反向排序正向排程（Backward Sequence Forward Schedule, BSFS）」的啟發式演算法，利用後向排序得到每個作業的容許等待時間再進行前向排程，同時判斷作業是否可以提前工作，藉此降低總延遲時間。BSFS的結果與Baker在1983提出的修正交期法（Modified Due Date , MDD）進行比較討論，試圖找出兩者之間對於總延遲時間的表現特性。依據本研究的實驗結果顯示，由於BSFS在逆向排程中是以交期為開工時間，而在進行派工時，則是以作業本身的作業時間為決策準則，故較不受交期大小而影響排程的結果，也因為加入詹森法則（Johnson’s rule）與『超車』概念所以在作業間的調整上較為靈活。

關鍵字

反向模擬

並列摘要

Quick and production runs is the trend of current manufacturing industry. How to reduce the total tardiness and minimize tardy jobs has become a major research topic and is critical to enterprise competency. There are many researches on dispatching rules of production scheduling proposing different solutions according to different situations. This thesis mainly discusses the schedule problem of Job Shop that takes into account due date and tolerant waiting time during scheduling. In order to achieve the goal, we have proposed a heuristic algorithm of Backward Sequence Forward Schedule (BSFS) which uses the backward sequence to obtain the tolerant waiting time for every operation and then proceeds the forward scheduling by making a judgment for a job and operation to be done earlier, in order to reduce the total tradiness. In this study, we also hold a comparison discussion with Bake’s Modified Due Date (MDD) proposed in 1983. From the comparison results we attempt to find out the performing characteristics regarding the total tardiness between BSFS and MDD. Because BSFS uses Due Date for ready time, during backward scheduling and it dispatching criteria is based on athe processing time of operation itself, due date has less effect on the scheduling according to our experimental results. Furthermore, adjustment among operations is more agile due to the adoption of Johnson’s rule and concept of cut-in.

並列關鍵字

Backward simulation

參考文獻

[1] Allahverdi, A. and Tatari, M.F., “Simulation of different rules in stochastic

[3] Kim, Y.D., “A backward approach in list scheduling algorithms for multi-machine

[5] Watson, E.F., Medeiros, D.J. and Sadowski, R.P. “Generating component release plans with backward simulation,” Proceedings of the 1993 Winter Simulation Conference, pp.930-938, 1995.

[6] Watson, E.F., Medeiros, D.J. and Sadowski, R.P. “A simulation-based backward planning approach for order-release, ” Proceedings of the 1997 Winter Simulation Conference, pp.765-772, 1997.

flowshops,” Computers & Industrial Engineering , Vol:31, Issue: 1-2, pp.209-212,