Comparing multivariable uncertain model structures for data-driven robust control: Visualization and application to a continuously variable transmission

The selection of uncertainty structures is an important aspect of system identification for robust control. The aim of this paper is to provide insight into uncertain multivariable systems for robust control. A unified method for visualizing model sets is developed by generating Bode plots of multivariable uncertain systems, both in magnitude and phase. In addition, these model sets are compared from the viewpoint of the control objective, allowing a quantitative analysis as well. An experimental case study on an automotive transmission application demonstrates these connections and confirms the importance of the developed framework for control applications. In addition, the experimental results provide new insights into the shape of associated model sets by using the presented visualization procedure. Both the theoretical and experimental results confirm that a recently developed robust-control-relevant uncertainty structure outperforms general dual-Youla-Kucera uncertainty, which in turn outperforms traditional uncertainty structures, including additive uncertainty.