本研究針對中心組裝廠評選產品零件之供應商問題進行探討。研究中指出於供應商遴選(Suppliers selection)時除應考量供應商之供貨狀況,亦應同時顧及組裝廠之生產情態。因此,本研究針對此問題建構一雙目標最佳化數學模式,並將組裝次序規劃(Assembly sequence planning, ASP)同步納入考量。同時,本研究採用非支配解排序基因演算法II(Non-dominated sorting genetic algorithm II, NSGA-II)與非支配解排序粒子群最佳化(Non-dominated sorting particle swarm optimization, NSPSO)進行模式求解,並發展混合型多目標演算法(Hybrid multi-objective algorithm, HMOA)。HMOA結合二種全域最佳化演算法,採用基因作業及粒子群速度與位置更新機制同步進行模式求解,流程中加入非支配解排序、排擠比較機制與利基法。一訂書機組裝案例被用以驗證HMOA之求解績效,評估結果顯示HMOA在精確性量測顯著優於NSGA-II與NSPSO,在分佈均勻性與延展性的量測比較上也較NSGA-II優異。
This study discussed parts supplier selection for central assembly plant. In addition to the delivery state of suppliers, the production conditions of the central assembly plant should be considered. A bi-objective optimal mathematical model was thus established for supplier selection in this study, which incorporates assembly sequence planning (ASP). At the same time, the hybrid multi-objective algorithm (HMOA) which combines Non-dominated sorting genetic algorithm II (NSGA-II) and Non-dominated sorting particle swarm optimization (NSPSO) was used to solve the model for a case of stapler assembly, in order to validate the solving effects of HMOA. The results showed that HMOA is not only more exact than NSGA-II and NSPSO, but also have a better distribution and a wider extent than NSGA-II.