Development of Production Simulator for Buffer Size Decisions in Complex Production Systems Using Genetic Algorithm

Abstract

Deciding on buffer size for production lines has gained more and more importance because of growing complexity of the production lines and production costs. Many researches have developed numerous techniques for gaining the optimal buffer size which can aid in achieving maximum production rate. One of the search methods that can be used for studying buffer size in production lines is a genetic algorithm (GA). In this paper, we propose a production simulator (PS) to determine optimal or near optimal buffer size and, at the same time, to minimize the number of generations required to reach this optimal buffer size for the complex production systems (CPS). Our PS consists of a GA and a discrete event. For the most efficient use of the GA we introduce a new encoding method for GA called a Multi-Vector Encoding Method (MVEM).