Summary: | 碩士 === 國立成功大學 === 工業與資訊管理學系碩博士班 === 100 === Scheduling is an important part of production management. Its purpose is to make full use of limited resource and to achieve maximum benefits for manufacturers. Currently, manufacturers are facing the problems regarding the ever-changing customer demands and product’s phase in and out generation after generation. Thus the concept of mass-produced products does not exist anymore.
Therefore, there will be higher chance that cost can be lowered and production process will be more efficient if some practical factors are taken into consideration. For example, shorter product life cycle will lead to shorter setup time. And the processing time can be shorten by applying appropriate training of employees. Furthermore, there is possibility that a best plan will be made and results in lower cost and higher efficiency if above factors are taken into account.
This research develops a single-machine scheduling model that involves the factors such as setup time, learning effect and time windows. The total punishment cost is the indicators of scheduling performance. The psuedo Branch and Bound Algorithm, Heuristic Algorithm and Genetic Algorithm are used in this research. Data simulation is used for testing the accuracy and computational efficiency of this algorithm. The results of the research are to help decision-makers making better decisions and provide reference for their further researches.
|