標題: 具順序相依整備時間與非固定序列特性之流程式製造單元排程
A Genetic Algorithm of Non-permutation Flow Shop Manufacturing Cell Scheduling with Sequence Dependent Setups
作者: 洪浩翰
Hung, Hau-Han
巫木誠
Wu, Muh-Cherng
工業工程與管理學系
關鍵字: 排程;流程式製造;基因演算法;染色體表達法;順序相依整備時間;genetic algorithm;chromosome representation,;scheduling, flowshop;manufacturing cell;non-permutation;sequence-dependent family setups
公開日期: 2009
摘要: 本論文是對於單元製造(Cellular Manufacturing)系統中,各個單元內的流程式製造單元(Flow Shop Manufacturing Cell)的排程問題。而研究的重點在於提出有別於過去的染色體表達法(Chromosome Representation),希望從染色體表達法的改善,以提高基因演算法的求解品質。探討的問題為靜態排程的架構,每個工件被劃分至不同的工件族(Family),而同一種工件族其特性為在同一機台加工,其間隔不需要額外的整備時間(Setuptime),而若是加工過程在同一部機台上,換成另一群的工件族來加工,而此時就需要整備時間,所以同一工件族的工件一起加工,加工完了才換至另一個工件族的工件加工,如此才可以達到最有效率的加工順序,而在流程式製造中,在機台與機台間如何排各工件族加工順序,以及各加工的工件族內,不同工件的加工順序,就是這篇論文所要探討的排程問題。 由於這個問題複雜度高,如果要使用整數規劃的方式求得最佳解,所需要花費的時間會而當高,所以本篇論文是以基因演算法來求解這個問題。 而本篇論文所提出新的染色體表達法以 GA-N 表示,而傳統的染色體表達法以GA-T表示。
This research is concerned with the scheduling of a non-permutation flowline manufacturing cell with sequence-dependent family setup times. We develop a new chromosome representation scheme. Based on the new representation and a prior one, two genetic algorithms are proposed to solve the scheduling problem. These two algorithms, respectively called GA-T (a prior algorithm) and GA-N (a proposed algorithm), adopt the same evolutionary mechanism but with different chromosome representation schemes. Numerical experiments indicate that GA-N outperforms GA-T in most test cases. This finding implies that developing new chromosome representation schemes might lead to better solutions quality for a particular meta-heuristic evolutionary mechanism. Choosing an appropriate chromosome representation is therefore very important to meta-heuristic algorithms.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT079733535
http://hdl.handle.net/11536/45442
Appears in Collections:Thesis


Files in This Item:

  1. 353501.pdf

If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.