Data Assimilation For Simulation-Based Real-Time Prediction/Analysis

There is growing interest in combining simulation and real-time data to support real-time prediction/analysis for complex dynamic systems. This paper presents a framework of using data assimilation to enable simulation-based real-time prediction/analysis for dynamic systems in operation. The different activities of the framework, including dynamic state estimation and online model calibration, are described. A demonstrative example is presented to show how these activities work together for a discrete event simulation application. Experiments results show the effectiveness of making data assimilation work with discrete simulations to support simulation-based real-time prediction/analysis.