Model order reduction for a full scale grid converter for grid simulations

Abstract

The simulation of transmission and distribution grids comes with a trade-off between the number of modeled components and the modeling depth on the one hand and the simulation speed on the other hand. An example system of a grid converter of a 2 MW full scale converter system is used to show a step-by-step model order reduction. Starting from a detailed switched model the order reduction is shown in two steps. First, the well-known averaging method is used to provide a time-continuous model and a time-continuous approximation of non-linear effects for the low-frequency range, that are usually lost due to averaging, are included. In a second step the model order reduction is continued based on the analytical equations of the continuous system. An order reduction method for linear systems is applied on a linear, time-invariant subsystem of the converter. The application of mathematical order reduction methods on the full average model reveals the limitation of these methods for the system in view, which stands in contrast to standard order reduction examples. An outlook is given on the application of advanced order reduction methods.