- 學位論文
半導體封裝壓模生產排程之研究
An Intelligent Scheduling Model for Molding Process in Semiconductor Assembly Industry
摘要
本文探討半導體封裝產業中壓模製程站之換模排程問題。壓模生產系統為平行機設備，壓模生產系統之所以能生產各式各樣不同種類之產品，完全取決於壓模模具之不同，才能生產各式各樣之產品。依據壓模站之產能限制，構建出一套換模模式，利用基因演算法，使固定週期內的投料產品及投料數量，求出滿足客戶訂單數量之換模模式，期望使訂單能順利生產，滿足客戶需求為目標。 本研究所研究的範圍，為研究彈性批量投料量之投料方式。而由於彈性批量投料量之投料方式，因此更機、換模成為無法避免的課題，現階段壓模機台模具之配備及安排，是人工依先進先出排程方式決定出來，生產線主管根據生管人員投料量，排定之生產換模計劃表，其中模具之換模時機及換模模具數量往往掌握不當或安排不當，造成換模次數增加。因此本研究之問題核心，在於透過有效的排程，改善先進先出人工排程不當之缺失。本研究結果，可以基因演算法之方式，找到當日最少之換模次數。
並列摘要
This thesis addressed a package change model of the assembly molding station in semiconductor packaging industry molding system is consisted of a se of a parallel machines. We use different mold dies to produce the products. The constrains of capacity of molding stations are considered in order to develop a suitable change package model. This study developed a genetic algorithm for finding the package change schedules which can minimize total change-over times. This thesis discussed range which was the flexible batch input model. The change over mold die to produce different product which is necessacy to be. Current mold die change which was depend on the production line supervisor use the FIFO ( first in first out ) rule to arrange. Base on the mold die capacity constrained, We use genetic algorithms for find package change model optimization solutions. To archive expected goal of facility which the shortest production time and the most less change mold die times.