摘要
With improved efficiencies, auto-guided vehicle (AGV) systems have developed rapidly and matured technologically in recent years. This research presents meta-model based two-stage simulation optimization for AGV systems with charging mechanisms. The optimal settings of the AGV system for the design and operational levels are proposed. Input data are collected from a real production environment. To develop the simulation model, several designs of the AGV system are developed and validated. This study aims to determine the number of required AGVs for the design-level simulation in the context of fixed guide-path constraints. For the operation-level issues, the charging system, positioning, and dispatching and routing rules for the AGV system are determined for the second-stage simulation. A factorial experiment is conducted using simulation models with different charging systems, in which output data are used to construct the metamodels. By applying global metamodel-based optimization, the obtained responses on the AGV utilization and throughput are optimized. Subsequently, a desirability approach is applied to optimize these responses. The proposed study is expected to provide a practical model for real application, facilitate the evaluation of policies prior to making decisions, and improve the efficiency of the AGV systems.