近年來隨著經濟環境與產業結構的快速變遷，市場上對於產品需求的大增，因此企業面臨訂單資源的不斷湧入，為了增加產能，企業不斷地擴廠或是合併其他產能，也使得原本簡單的訂單管理問題，隨著工廠的增加、地域的拉遠，從單廠衍生到複雜的多廠規劃(Multi-Site Planning)問題上。在本研究中，將針對多製造廠在全面訂單管理的情況下，探討在一健全的訂單管理制度下，對於製造廠的跨廠訂單分配模式進行研究，建立一個考慮到多製造廠在接收大筆訂單時，如何依據產品的市場銷售特性、訂單交期、廠與廠間的生產排程、訂單利潤及廠的產能負荷度等等交叉複雜的關係，建立一多廠整合型生產指派與排程系統的決策模式。而現實世界的排程問題並非單一目標可以滿足，需以多目標(multi-objective)的觀點來考慮，而且目標之間每每是彼此衝突。而目前多數關於多目標排程的研究中，僅考慮生產製造方面的績效因子，這些因子多是屬於量性因素(quantitative factors)。事實上，排程環境中所需考量的還需包含有一些關於組織營運策略上的質性因素(qualitative factors)。本研究以混合式遺傳演算法(Hybrid-GA, HGA)求算訂單指派至各廠的生產排程與各廠的產能平衡，以決定訂單的最佳配置。進而便能依據所指派的結果作為各家製造廠進行生產計畫與排程規劃的依據。整體來說，本研究擬提出一套多廠整合型生產指派與排程系統，最後透過Job shop排程實證問題，驗證本研究所提方法與求解模式的可行性及分析衡量此系統帶來的效益。 As economic environment and industrial structure changing rapidly, product demand frequency increases immensely as well. To accomplish the great amount of orders, industries expand their plants accordingly or integrate with others to increase their outputs. As a result, conventional order management is no longer sufficient to solve these problems, accompanied with the distance and different information of these incorporated factories. The single-site planning has developed in to multi-site planning.This research aims to construct the decision model of integrated multi-site production assignment and scheduling problem. To support the multi-site factories with their enormous orders, based on the premise that they are under total order management, the decision model considers such complicated factors as product market features, due date, production schedule, order profit and capacity load of each plant.There is no more a singular objective in real world scheduling system, but multi objectives that are commonly conflicting to each other. In addition, the effect factors taken account by current multi-objective scheduling research are quantitative factors. Essentially there are more qualitative factors to be considered related to organizations’ operating tactics.Our research using hybrid-genetic algorithms method for production assignment’s scheduling and each plant’s capacity balancing to determine optimal order’s allocation. Consequently, our research proposes an optimization of multi-site integrated production assignment and scheduling system.
