Augmented Lagrange Multiplier Functions and Duality in Nonconvex Programming

If a nonlinear programming problem is analyzed in terms of its ordinary Lagrangian function, there is usually a duality gap, unless the objective and constraint functions are convex. It is shown he...