近年來由於科技日新月異,工廠的生產趨向於快速生產,在整個生產過程中,研究指出製造成本約有13%?30%是來自於物料搬運成本,因此,選擇一個適當的物料搬運系統,對於成本的降低是具有顯著的影響。在許多的物料搬運系統中,無人搬運車是一具有彈性與生產力的物料搬運系統,由於無人搬運車的彈性途程特性,使得在工業界的應用與學術上的研究非常地廣泛。其相關的研究範圍包括:軌道設計、指派問題、車輛需求評估、與車輛防撞、車輛承載問題以及在不同的環境下車輛的績效評估等問題。而本研究是在一個單載量無人搬運車及JIT的系統環境下,發展一在封閉式系統中,以懲罰值為基準的AGV控制方法,藉此處理無人搬運車的指派問題。 本研究將使用ARENA 9.0(SIMAN語言)模擬軟體來撰寫模擬程式,利用模擬程式來架構出所有的環境假設及單載量無人搬運車系統在不同以懲罰值為基準的載取派車法則之模擬實驗並且和最大需求(MD)法則做一比較。在本研究中,車輛控制法則是由3種AGV行駛速度、3種不同系統WIP情況及4種載取派車法則設定所搭配出108種不同車輛控制情境,本研究將比較這36種車輛控制情境,根據工件平均流程時間、工件平均產出量、提早工件之平均提早完成時間、延遲工件之平均延遲完成時間、工件之平均提早延遲完成時間及總平均Lateness等6個不同績效衝量指標,藉由實驗設計過程和統計分析來了解各法則的績效與差異。 In a production process, about 13% to 30% of production cost is from material handling cost. Therefore, selecting an appropriate material handling system is critical to reduce the production cost. Automated guided vehicle (AGV) is a material handling which has flexibility and productivity. Because of the flexible routine of the AGV, there are many industrial applications and academic researches in recent years. This research is focused on a JIT environment where is closed system and developing a single load AGV dispatching rule based on penalty value to process the material handling problems. We will use Arena 9.0 to simulate all of our experiments. There are three main factors in our experiments, there different AGV velocity, there different WIP quantity and four dispatching rules. And we use six different performances (average flow time, average throughputs, average earliness plus tardiness, average lateness, average earliness and average tardiness) to analyze the performance and difference of different AGV dispatching rules.