Model Reduction and Assessment for Nonlinear Networked Systems
Friday, September 17, 1999 - 10:10am - 10:50am
Reduced models are an important tool for simulation and control of large-scale nonlinear dynamical systems. We outline some issues in determining an effective reduced model, and give examples of model reduction techniques applied to nonlinear networked systems from chemical kinetics and from power systems. For highly nonlinear systems, the reduced model can be very data dependent. We investigate several means of assessing the accuracy of a reduced model over a range of data.