面對日益複雜的生產系統,工廠資源衝突亦漸趨惡化,如何將有限的生產資源做最適化搭配,以配合顧客需求及追求公司利潤最大化,是產業界極為重視的一項問題。在半導體晶片最終測試產業中,顧客產品必須配合測試機台與其他輔助設備和資源相搭配,才得以遂行晶片測試工作;亦即,在公司多重資源限制下,進行產品的測試作業規劃。鑑於此一產能分配的問題十分重要,本研究提出在同時考量公司資源(如測試機、送料器等)的產能分派模型。 本論文以某半導體測試產業為例,探討一規劃期間(Planning horizon)內,已知顧客需求(產品組合)之產能分派問題。考慮多重生產資源限制下,以追求公司利潤最大化或最小總加工時間為目標,進行公司產能分派。本研究以限制滿足規劃法建立產能規劃模型,並以限制為基之推理引擎應用程式為工具,實作並驗證此一多資源與資金限制下之產能分派決策輔助系統。
As the complexity of production system increases, it is serious in the production resources conflicting environment. It has become the issue as to how to allocate the production resources appropriately in order to satisfy the customer demands and gain the maximum profits. In semiconductor final test industry, the products need to use various production resources so that testing processes have the multi-resource constraints in capacity planning. This research focuses on the issues pertaining to (i) what kinds of customer orders to receive under multi-resource constraints (ii) how to allocate tester and handler capacity under different objective functions so that customer’s demand can be satisfied. We developed a capacity allocation model for the two issues that take into consideration in multi-resources constraints of the final test processes at the same time. This study uses constraint-based development tool to develop this model and uses constraint-based search methods for solving this problem. The experiment results show how to allocate tester and handler capacity and the utilizations of tester and handler. Capacity allocation can help to make the decisions in capacity planning.