Masters' Theses A hybrid algorithm for production scheduling under Make-to-Order environment : a case study from the transformer industry
2015.08.04 17:36
| year | 2014 |
|---|---|
| author | Sung Bum Jun |
| Keyword | Hybrid flow shop 스케줄링 문제, 동시 작업, 야근 작업, 변압기, 주문형 생산 시스템, 혼합 알고리즘, Hybrid flow shop schedulingproblem, Make-to-Order production system, Transformer industry, Nighttime work, Simultaneous work, Hybrid genetic algorithm |
| Abstract | The industrial transformer market has a characteristic of flow shop production system by specific requirements such as due date, voltage, and capacity. In this Make-to-Order production environment, on-time delivery has a major influence on the competitiveness of company. From the manufacturing factory’s standpoint, it is also important to minimize the production cost like work-in-process inventory holding cost and transporting cost between stages by the due date that a company promised their clients. Therefore deriving a schedule to minimize both tardiness and production cost is essential for improving business competition. However the complexity of flow shop problem and a lot of constraints in the real world make it harder to solve these problems within a reasonable time. This paper presents a hybrid flow shop scheduling problem with real-world constraints, and develops a hybrid genetic algorithm for its solution. We first discuss the characteristics of the hybrid flow shop problem under the constraints of nighttime and simultaneous work. A hybrid genetic algorithm is then formulated to minimize the total tardiness. This algorithm incorporates both Nawaz– Enscore–Ham (NEH) and local search algorithms. The performance of our proposed approach with those of heuristic algorithms is compared. The results show that the proposed algorithm outperforms the NEH algorithm, a simple genetic algorithm, and five existing dispatching rules in terms of a total tardiness performance measures. |
| c | MS |
Download :
