A selection approach for solving buffer allocation problem

Abstract

In this paper, we dealt with one problem in designing a production line, which is the problem of buffer allocation. A selection approach was developed and tested for selecting the best design for a huge number of alternatives set. The proposed selection approach is a combination between cardinal and ordinal optimization. The algorithm involves four procedures; ordinal optimization, optimal computing budget allocation, subset selection and indifference-zone. The purpose of this paper is to use the proposed selection approach to find the optimal allocation of buffers that maximizes the mean production rate (throughput) in short, unbalanced and reliable production lines. Numerical results are presented to demonstrate the efficiency of the selection algorithm in finding the best buffer profile where its mean production rate is at its maximum. Key words: Buffer allocation problem, production lines, ranking and selection, ordinal optimization, optimal computing budget allocation, optimization.