This paper presents a methodology to optimize production rate and focuses on an optimal buffer storage sizing for this purpose. Buffer storages are intermediate semi complete components which their capacity is an important factor in design of production lines and impact on the production line. This methodology with eliminating of some simpler assumptions, assumes that the arrival rate of parts to production (or assembly) line, service times of the stations, time between failures of machines and repair time of failed machines are non deterministic and generally distributed (not just exponential). To optimize the buffer capacities, we consider production line as a queuing network and we use response surface methodology, simulation, genetic algorithm and multi objective decision making

Keywords: Buffer storage, Queuing network, Genetic algorithm, Response surface methodology, Simulation, Multi objective decision making

Full-Text [PDF 4286 kb]   (3150 Downloads)