An inexact logarithmic-quadratic proximal augmented Lagrangian method for a class of constrained variational inequalities (Q2472183)

The authors present an augmented Lagrangian method for the solution of constrained variational inequalities according to \[ x\in S,\;(x'-x)^T g(x)\geq 0,\quad\forall x'\in S\quad\text{with }\;S= \{x\in\mathbb{R}^n\mid Ax= b,\;x\geq 0\}, \] where \(A\in\mathbb{R}^{m\times n}\) is a given matrix, \(b\in\mathbb{R}^n\) is a given vector and \(g: \mathbb{R}^n_+\to \mathbb{R}^n\) is a given monotone mapping. By introducing special logarithmic-quadratic proximal terms, the nonlinear complementarity problems in each iteration step of this method are transformed to systems of nonlinear equations (with unique solution) which are solved approximately under a rather relaxed inexactness criterion. The global convergence of the generated sequence is proved. Important implementation details and some numerical results for special problems are pointed out at the end of the paper.