Capacity planning by the dynamic multi-resource generalized assignment problem

The allocation of available capacity among competing demand and users is a problem encountered in areas such as job shop scheduling, the trucking industry and distributed computer systems. In all these areas a model known as the Multi-Resource Generalized Assignment Problem (MRGAP) has been proposed as a tool to assign available capacity among the competing applications. In this paper we extend the MRGAP model to the case where demand varies over time and capacity assignments are dynamic. We show that the extended model can be used for strategic capacity planning and we develop efficient solution procedures to solve the dynamic version of MRGAP.