Volume 2, Number 3, September 2006, pp.657-665
	Rolando Gárciga Otero
				Key words:
			augmented Lagrangian, cone-constrained optimization, inexact solutions, strong convergence
		Mathematices Subject Classification: 90C25, 90C30, 49J40, 46M37

ONLINE SUBSCRIPTION (Institutional Subscription Only) Copyright© 2006 Yokohama Publishers Back

---

	Abstract:
	This paper deals with the general optimization problemmin g(x) subject to -G(x)∈ K, with g: X → R, G: X → Y, where X and Y are real reflexive Banach spaces and K is a nonempty closed convex cone in Y. An augmented Lagrangian method is proposed for this problem, which allows for inexact solutions of the primal subproblems and guarantees strong convergence of the primal-dual sequence of iterates to an optimal pair. Moreover, the relation between the initial primal-dual iterate and the strong limit is established.

A strongly convergent augmented Lagrangian method

---