Middleware Solutions for Next Generation Integrated Simulation Environments.
详细信息
文摘
Simulation and Modeling is one of the most important and widely used techniques in many research areas. Simulation is particularly indispensable for studying large-scale and complex systems,which can be otherwise intractable to closed-form mathematical or analytical solutions. One of the limitations of simulation models is that they are typically designed to be executed and evaluated independently. The available simulation tools are mostly meant for standalone use and do not attempt to address the overall scenarios. It would be advantageous and cost-effective to combine these models thus forming a new integrated simulation model based on individual existing components to model complex scenarios and understand the joint effect of multiple phenomena. This will allow decision makers to explore alternatives and conduct low cost experiments in an integrated environment. The integration of different simulation models can not be done easily,since each of the individual models are implemented in part with different simulators which do not interoperate with one another. We need to integrate different expert simulation models that each captures relevant aspects of the real world. In this thesis,we propose a framework to integrate multiple autonomous simulation models into an integrated environment wherein we can model and execute complex scenarios involving multiple simulators. Our techniques are implemented in RAISE Reflective Architecture for Integrated Simulation Environments)； specifically,we design meta-level synchronizer and data exchange modules that support flexible constraint specification while supporting high levels of concurrency in the execution. We evaluate our proposed techniques via a detailed case study from the emergency response domain by integrating different real-world simulators &ndash； a fire simulator CFAST),an evacuation simulator Drillsim),a transportation simulator TranSim),and a communication simulator LTEsim). We believe that our proposed techniques will serve as fundamental tools and building blocks for building the next generation integrated simulation environments.