Convergence properties of modified and partially-augmented Lagrangian methods for mathematical programs with complementarity constraints (Q983712): Difference between revisions

Mathematical programs with complementarity constraints (MPCC) form an important class of mathematical programs with equilibrium constraints and have wide applications in economics and engineering. Due to the presence of the complementarity conditions, the Mangasarian-Fromovitz constraint qualification and the linear independence constraint qualification are not satisfied at a feasible point of MPCC. As a result, the traditional nonlinear programming methods cannot be applied directly to MPCC. Complementarity problems can be considered as special case of variational inequalities. In recent years, much attention has been drawn to the study of MPCC. The purpose of this paper is to investigate the convergence of partially augmented Lagrangian methods for solving MPCC without appealing to the restrictive boundedness assumption of the Lagrangian multipliers. Four modified partially augmented Lagrangian methods for solving MPCC based on different algorithmic strategies are proposed and analyzed. The convergence to a B-stationary point of MPCC is established for the four modified partially augmented Lagrangian methods without requiring the boundedness of the Lagrangian multipliers.