Optimal Scheduling of Automatic Guided Vehicle System via State Space Realization

Abstract

This paper considers dynamical system modeling of transportation systems in semiconductor manufacturing based on state space realization. Utilizing this method, we consider an optimal scheduling problem for an Automatic Guided Vehicle (AGV) transfer problem, which is to control AGV congestion at transport rail junctions. Our scheduling algorithm is based on model-predictive control in which the cycle of measurement, prediction and optimization is repeated. Its optimization is recast as an Integer Linear Programming (ILP) problem. Since little attention has been given to AGV scheduling based on model-predictive control, no method is, to our knowledge, known for determining appropriate cost functions. Here, we focus on throughput maximization and shortest transit time problems and show corresponding cost function settings. We also propose a visualization algorithm of AGV scheduling via state space realization, presenting numerical examples.