薄膜液晶顯示器產業多階多廠之生產規劃機制

Abstract

TFT-LCD產業包含陣列、組立及模組三大主要製程，其中每個製程包含多個不同世代之單廠，造成其多階多廠之生產環境。除此之外，整個產業又具有許多特有之生管特性，例如對各製程之完成品作產品分級制度、組立製程每二天之固定整備一次、各製程良率不同以及陣列階段不同世代廠之玻璃基板大小不同所造成面板產出量不一等特性。因此本文發展出一適宜之生產規劃機制以規劃出各廠最適宜生產的產品別及數量來滿足各產品別在各階段之需求同時極大化整體之利潤。 本文以啟發式之「需求規劃模組」及「產能分配模組」建構出TFT-LCD產業之多階多廠之生產規劃機制。首先，「需求規劃模組」以「產品等級別需求量推估流程」，推算出製程內各產品之各等級別之毛需求量。接著，以所求得之各產品之毛需求量及存貨資訊，考量製程良率及不同世代廠玻璃基板之面板切割數，以「製程別淨需求量推估流程」，求得各製程階段之產品別淨需求量。最後，考量各製程各階段各廠各世代的產能限制，以混合整數規劃建構「陣列製程產能分配模式」及「組立製程產能分配模式」，以求解各製程階段各廠在各規劃週期內所應生產的產品種類及數量。 模擬結果顯示，本文所提之生產規劃機制確實可以滿足各種最終產品之需求，並可達到良好之訂單達交率及利潤極大化之目標。

The three main processes of TFT-LCD industry are array, cell and module. Each process stage contains various generation plants with different production capability such that a multi-echelon environment is composed. Besides, there are many particular production characteristics inherited in the whole process, such as grade setting of each finished goods at each process, fixed setup occurred every two days in cell process, various yield rates in each process, and a glass panel from different generation of array process scribed into different pieces of LCD panels. Thus, in this thesis, we established a production planning mechanism so as to decide proper product throughput of each plant to fulfill their demand needs respectively as well as to achieve maximization of profits entirely. Our proposed production planning mechanism composed of a heuristic “demand planning” module and a “capacity allocation” module. We first calculate the gross demand for each product grade by using “gross demand derivational process with grade classification”. Next, according to information of gross demand, inventory, yield rates and panel conversion rate between array process and cell process, we calculate the net demand requirements for each process through “net demand derivational process”. Finally, we provided a mixed integer programming model as the “capacity allocation model” so as to decide the production quantity at each plant in each planning period. Simulation results reveal that by using our proposed planning mechanism, we achieve high fulfillment rate for every final product type as well as maximize profits of whole enterprise simultaneously.