以基因演算法求解彈性製造工廠之製程規劃與排程問題

Abstract

本研究探討一個建立在彈性零工式排程 (flexible job shop scheduling, FJS) 之製程規劃與排程整合 (integrated process planning and scheduling, IPPS) 問題。彈性製造工廠之製程規劃與排程問題的決策變數有三個，分別為：(1) 製程規劃指派：工件指派到哪一個製程規劃，(2) 作業指派機台：作業指派到哪一個機台上加工，(3) 作業排序：機台上的各個作業如何排序。 製程規劃與排程整合問題意指將製程規劃與排程同時進行決策，以往製程規劃與排程相互獨立進行決策，導致生產效率不佳，為對此問題進行決策，本研究提出以工件指定之製程規劃機率值，及以工件表達的作業順序所組成的兩段式基因染色體表達法，並發展基因演算法求解製程規劃與排程整合問題，不只將製程規劃與排程進行整合，解決資源上的衝突，更提升生產力。透過與Amin-Naseri et al. (2012) 進行實驗比較，實驗結果顯示，本研究所提出之新基因染色體表達法，並發展基因演算法，有91%的機率能夠有效解出彈性製造工廠之製程規劃與排程問題。

This research studies an integrated process planning and scheduling (IPPS) problem in the context of a flexible job shop (FJS). An IPPS problem denotes that the process planning problem and scheduling problem are simultaneously solved. An FJS is a job shop in which an operation can be assigned to multiple machines. The problem is herein called an IPPS-FJS problem, which includes three decisions: (1) job-to-process plan assignment, (2) operation-to-machine assignment, (3) operation-sequencing. This study proposes a new chromosome representation and proceeds to develop a genetic algorithm (GA) to solve the IPPS-FJS problem. The work of Amin-Naseri et al. (2012), the state-of-the-art in solving IPPS-FJS problem, is taken as the benchmark for justify the effectiveness of the proposed GA. Experiments reveal that the proposed GA outperforms the benchmark algorithm.