A comparative Study of Model-Based and Data-Based Model Order Reduction Techniques for Nonlinear Systems

In this paper a comparative study of three nonlinear model order reduction techniques using a case study of a transmission line model is presented. The investigated model order reduction techniques are: quadratic approximation, trajectory piecewise linear approximation and data-based identification of bilinear model. The performance of model order reduction techniques has been evaluated in terms of their accuracy and computational cost. The original 100 th order nonlinear model is reduced to 12 th and 20 th order models by using three different MOR techniques yet preserving simulation accuracy.