本研究探討整合單一製造商與單一產品含機台隨機當機之(n+1)次配送策略的最佳生產時間。研究中假設在不完美的生產過程中會有不良品的產出,其中有一部分的不良品可進行重工修復,其餘無法修復之不良品即為報廢品,而重工過程中無法修復的則視為報廢品。為了滿足顧客需求,因此在生產時間內先進行一次的配送,待重工修復後進行n次的配送。在生產的過程中機台會有隨機性的當機情況發生,而這個隨機當機的情況可能會發生在第一次配送前或第一次配送後的生產時間,而機台發生當機的情況時則採用AR(abort/resume)策略作為機台維修後的處理方式,即當機台維修完成後繼續生產未完成的批量。 本研究針對上述假設建立了三種可能發生的生產模型:(1)當機的情況發生在第一次配送前的生產時間內;(2)當機的狀況發生在第一次配送後的生產時間內;(3)生產的過程沒有發生當機的狀況。針對這三種可能發生的生產模型分別建構出各自的成本結構,加以推導,利用積分的方式整合這三個數學模式以求得最佳之生產時間使期望總成本最小化,最後提出數值例子來加以驗證與敏感度分析,以提供業界作為生產決策之參考。
This study determines the production run time for a manufacturing system with stochastic breakdown, (n+1) delivery policy, and partially reworking of defective items. The classic Economic Production Quantity (EPQ) model assumes the manufacturing system is perfect during the production run time. But, in real-life manufacturing system, defective items and stochastic machine breakdown are inevitable. In this research, we assumed that the manufacturing system produce defective items randomly, a portion of defective item can be repaired through rework, the others are considered scrap. Furthermore, the manufacturing system is subject to stochastic machine breakdown, and when it occurs the AR policy is adapted. We assumed that one upfront distribution of finished items is shipped to meet customer’s demand within the production and rework time, then in the end of rework, multiple distributions are done. Mathematical modeling and analysis are used to solve this model. Numerical example and sensitivity analysis is provided to demonstrate its practical usage.