本文以塑膠產業中的拉門製造廠之排程優化為研究主題，發展可選擇績效指 標下以改良式基因演算法為基礎之排程演算法。考慮在多工作站、批量生產且不 可插單的限制下，尋找最佳排程規劃，以達總完成時間最小、總延遲時間最小或 總延遲訂單數最少之目標。本研究所提出之改良式基因演算法，以輪轉編碼及交 配池為基礎，以坊間容易取得之軟體EXCEL2013及Visual Basic2010作為開發平 台，並與原工廠之排程方法進行比較。結果顯示本研究所提出之改良式基因演算 法，在各項績效指標下，皆優於原工廠所使用之排程方法。

This article focuses on scheduling optimization for a door manufacturing company, based on various performance indicators; a modified genetic algorithm method is developed to search for the optimal production scheduling. In this paper, multiworkstation, batch production and order interfere prohibit are considered. The performance indicators include: 1. Minimum the total makespan, 2. Minimum the total delay time, 3. Minimum the total number of delay orders. The software platforms of the proposed scheduling system are EXCEL2013 and Visual Basic2010. The modified genetic algorithm method is used with rotary encoder and mating pool. Based on proposed method, the results are more efficiency than the original scheduling method under any performance indicator.