Optimization of buffer allocation in unreliable production lines based on availability evaluation

Summary The buffer allocation problem is an important issue in production lines design. In this paper, we present new evaluation and optimization methods to optimally allocate buffers in unreliable production lines. Through analyzing different states of the machines and buffers by Markov process and incorporating the aggregation method, we make an evaluation on the system availability, instead of the throughput rate of the line. The optimization method is proposed by combining particle swarm optimization and estimation of distribution algorithm to maximize the system availability. It generates the new populations by estimation of distribution algorithm and particle swarm optimization to take their respective advantages in global and local optimization. Numerical tests and simulations are performed to validate the performance of the evaluation and optimization methods. The results indicate the effectiveness and efficiency of the proposed methods.