Model composability and execution across simulation, optimization, and forecast models

We present a novel simulation platform called Optimization, Simulation, and Forecasting (OSF) for the domain of manufacturing and logistics supply-chain systems. It supports composition of DEVS, Linear Program (LP), and forecast models using an extended Knowledge Interchange Broker (KIB). Models developed in DEVS-Suite simulator, OPL-Studio optimization, and a heuristic Inventory Strategy Forecaster (ISF) can be composed using a new set of scalable XML Schemas developed for DEVS, LP, and ISF models. The addition of forecast modeling offers new kinds of supply-chain system simulation. In particular, alternative customer demand forecast “bias correction” methods can be evaluated towards optimized operation of supply-chain processes. The OSF platform affords modeling interactions among process, optimization, and forecast models. The KIB coordinates simulation (execution) of the DEVS, LP, and ISF models in a sequential fashion. Composition of each pair of DEVS, LP, and ISF models leads to scalability for specifying model interactions. Independent execution of each model allows flexible computation platforms. They simplify defining a large number of different data transformations. The concept, basic architectural design, and implementation of this composable simulation platform are highlighted using example single-echelon and multi-echelon semiconductor manufacturing, logistics systems.