Performance Optimization of Production Lines Via Buffer Allocation With Adaptive Genetic Algorithm

Abstract

The buffer allocation problem (BAP) is one of the most important problems in production line design and optimization. Under the constraint of total buffer size, find the optimal buffer configuration to maximize the Production Rate (PR) of unreliable production lines. BAP is an NP-hard combinatorial optimization problem, and its solution space grows exponentially with the size of the problem. Therefore, metaheuristic algorithms are widely used to solve BAP. In this study, we propose a hybrid adaptive genetic algorithm (AGA) and simulation-based approach to solve BAP, using a simulation model that simulates production line behavior to evaluate the applicability of each solution. Conduct numerical experiments on existing benchmark problems for unreliable production lines of different scales. The proposed method is compared against scenarios with no buffers, average buffer allocation, and existing methods from the literature. The purpose is to demonstrate the effectiveness of the optimal buffer configuration. For the four types of production lines, the PR of the proposed method are 0.87695, 0.50860, 0.65171, and 0.70492, respectively, which are 9.75%, 40.76%, 121.55%, and 137.92% higher than those without buffer zones, and are superior to the three benchmark methods. It can also enhance the utilization rate of the device, with improvement effects of 9.45%, 49.44%, 127.87%, and 140.14%, respectively. The AGA can shorten the optimization time, making up for the shortcomings of simulation methods. The simulation results show that our method is effective in maximizing PR.