Practical Stabilization of Switched Affine Systems: Model and Data-Driven Conditions

Abstract

An important trend about data-driven systems has recently emerged motivated for practical issues, that are the difficulties to identify a model of the plant. In this direction some recent results enhanced the robust control theory to get efficient and relevant solutions. To the best of or knowledge, the existing solutions are rather limited to the linear case, and only few addresses the stabilization of nonlinear systems, which still represents a major challenge. In this work, the objective is the design of switching control law for switched affine systems, ensuring that a neighborhood of a given operating point is stable. After proposing a model-based control design, we propose a systematic method to translate a model-based condition expressed using the framework of Linear Matrix Inequalities (LMI), into a data-driven condition. This is made possible thanks to a new formulation of the existing method from the recent literature on this topic. The method is then evaluated on an academic example.