On the Linear and Asymptotically Superlinear Convergence Rates of the Augmented Lagrangian Method with a Practical Relative Error Criterion

TL;DR

This paper analyzes the convergence rates of the augmented Lagrangian method with a practical error criterion, showing conditions for linear and superlinear convergence depending on the penalty parameter.

Contribution

It provides a detailed convergence rate analysis under mild conditions, revealing how the penalty parameter influences the rate, including superlinear convergence with increasing penalty.

Findings

Locally Q-linear convergence under mild error bounds

Superlinear convergence as penalty parameter increases to infinity

Convergence of the primal sequence distance to the solution set

Abstract

In this paper, we conduct a convergence rate analysis of the augmented Lagrangian method with a practical relative error criterion designed in Eckstein and Silva [Math. Program., 141, 319--348 (2013)] for convex nonlinear programming problems. We show that under a mild local error bound condition, this method admits locally a Q-linear rate of convergence. More importantly, we show that the modulus of the convergence rate is inversely proportional to the penalty parameter. That is, an asymptotically superlinear convergence is obtained if the penalty parameter used in the algorithm is increasing to infinity, or an arbitrarily Q-linear rate of convergence can be guaranteed if the penalty parameter is fixed but it is sufficiently large. Besides, as a byproduct, the convergence, as well as the convergence rate, of the distance from the primal sequence to the solution set of the problem is obtained.