An improved linearization point selection method in trajectory piecewise-linear model order reduction

The trajectory piecewise-linear (TPWL) method is a promising model order reduction method for nonlinear systems. Since the quality of linearization points has an important influence on the accuracy and complexity of the obtained reduced-order model, how to select them reasonably is crucially important in TPWL method. In our previous study, we have presented a linearization point selection method based on a global maximum error controller. It has been shown that this method has many advantages over other existing methods. In this paper, an improved version of this method is presented, which introduces a preliminary linearization point selection procedure. Compared with the original method, the improved method takes less time to extract a TPWL model of similar quality.