Extrapolated Proportional-Integral Projected Gradient Method for Conic Optimization

TL;DR

This paper introduces xPIPG, a new first-order method for conic optimization that detects infeasibility and outperforms existing solvers on large-scale problems, with applications in model predictive control.

Contribution

The paper presents xPIPG, a novel extrapolated proportional-integral projected gradient method that automatically detects infeasibility and improves convergence in conic optimization.

Findings

xPIPG effectively detects primal or dual infeasibility.

xPIPG outperforms existing solvers on large-scale problems.

Demonstrated success in model predictive control applications.

Abstract

Conic optimization is the minimization of a convex quadratic function subject to conic constraints. We introduce a novel first-order method for conic optimization, named \emph{extrapolated proportional-integral projected gradient method (xPIPG)}, that automatically detects infeasibility. The iterates of xPIPG either asymptotically satisfy a set of primal-dual optimality conditions, or generate a proof of primal or dual infeasibility. We demonstrate the application of xPIPG using benchmark problems in model predictive control. xPIPG outperforms many state-of-the-art conic optimization solvers, especially when solving large-scale problems.