In this letter, we investigate sufficient conditions for the exponential stability of LTI systems driven by controllers derived from parametric optimization problems. Our primary focus is on parametric projection controllers, namely parametric programs whose objective function is the squared distance to a nominal controller. Leveraging the virtual system method of analysis and a novel contractivity result for Lur'e systems, we establish a sufficient LMI condition for the exponential stability of an LTI system with a parametric projection-based controller. Separately, we prove additional results for single-integrator systems. Finally, we apply our results to state-dependent saturated control systems and control barrier function-based control and provide numerical simulations.
@article{arxiv.2403.08159,
title = {Exponential Stability of Parametric Optimization-Based Controllers via Lur'e Contractivity},
author = {Alexander Davydov and Francesco Bullo},
journal= {arXiv preprint arXiv:2403.08159},
year = {2025}
}