Abstract
工作分派不僅在生產線設計階段是一個重要的考慮主題，面對多樣化之顧 客需求與訂單大小，更能藉由適當的分派法則提高生產效率。本研究之主 要目的就是在因應不同之生產型態與需求之下，發展出一套完整之工作合 併與分派法則。從管理的角度來看，由於人員工作之不確定性，減少人員 之雇用是每一個管理者的原則。在人員派工方面，已發展出許多法則來解 決一些特定的系統，然而就人員精簡的考量，卻較少研究。我們為減少人 員指派並發揮最大之生產力，提出一套兩階段式的分派法則，來解決工作 分派的問題。當然，再進行分派之前，我們也整理出一套工作分析、歸類 、與量測之準則。為了能有效地應用這些方法，我們將這些法則用 Turbo PASCAL 語言發展成為實用的程式，以使現場管理工作更具彈性和 效率。程式的執行結果以商用之模擬軟體 STARCELL 來驗證。在案例分析 與應用中，我們可以清楚地看出這套派工法則的應用將不僅限於生產線之 工作，並可以成功地應用到其他行業，如運輸業、服務業等。 Job assignment is an important topic in the design of a production line. It can affect the productivity of a line especially in the face of a changing demand quantities. The primary purpose of this study is to investigate the optimal job assignment under the condition of different throughput requirements and resource constraints. From the manager's point of view, it takes more time and costs to deal with operators than with machines due to the uncertainty of human operations. Therefore, most managers would like to hire the minimum number of operators.Previous researchers on the worker assignment problems have developed some job assignment policies, but few of them pay attentions to minimize the number of workers. In this research, a two stage methodology will be used to improve the productivity of a line with some resource constraints. Before applying this methodology, the basic tasks have to be identified and their processing times have to be measured. Furthermore, the tasks must be classified into different types based on the degree of operator involvement. In order to make the best use of the proposed methods, it is necessary to develop some computer programs for the applications. The algorithms are programmed in the Turbo PASCAL language. Moreover, the performance of the proposed methods will be verified by a commercialized simulaion software STARCELL. In addtion to the design of production lines, these algorithms can also be applied to other fields. Some cases of production systems and transportation systems are illustrated in this study to show the applications of these methods.