Model reduction preserving the steady-state gain and its application to electric power systems

摘要

This paper studies a model reduction method via balanced truncation. In order to reduce the approximation errors in step responses, we consider a new method that preserves the steady-state gain in the model reduction. This method is especially effective for strictly proper systems. We intend at the same time to keep the H{sub}∞ norm of the error system as small as possible, and the method employs the LMI optimization technique. The mothod is applied to the reduction of an electric power system as an example.