이산사건 시뮬레이션과 유전자 알고리즘을 이용한 제조업 공장의 라인 최적화

Abstract

This paper proposes the application of the genetic algorithm to the discrete event simulation of manufacturing lines to find the optimal set of the line parameters for line balancing. The proposed method was applied to two example problems, and thus was shown to be useful for line balancing. In spite of the rapidly increasing interest in digital manufacturing, it still lacks a systematic approach. In particular, the line flow analysis method via modeling and simulation is highly demanded; yet, when the line is to be improved, the new line parameters are fixed by engineers solely based on their experience. The method proposed in this paper enables to automatically and objectively search an optimal set of line parameters for best line balancing. Among various optimization methods, the genetic algorithm was adopted because it is known to be robust and suitable for the discrete event space. DELMIA V5 QUEST of Dassault Systems was used for the discrete event simulations, and the genetic algorithm was written in the Simulation Control Languages within QUEST. The developed system was applied to a very simple model and a real model of shipyard, and its validity was confirmed, and its usefulness demonstrated.