We study the problem of learning-augmented predictive linear quadratic control. Our goal is to design a controller that balances consistency, which measures competitive ratio when predictions are accurate, and robustness, bounds inaccurate.

<p style='text-indent:20px;'>In this paper, an optimal control model ruled by a class of linear discrete-time stochastic descriptor systems is considered under quadratic index performance. Employing dynamic programming method, recurrence equation to simplify the problem presented provided that are both regular and impulse-free. When objective function quadratic, according equation, linear...

This technical article investigates the linear quadratic regulator (LQR) design for continuous-time positive systems. Based on systems theory and Lyapunov theory, solvability optimality of positivity-preserving LQR problem are analyzed through lens optimization, two projection theorems derived single-input multi-input systems, respectively, which paves way developing a projected gradient descen...