1: \begin{abstract}

2: This paper extends the SQP-approach of the well-known bundle-Newton method for nonsmooth unconstrained minimization to the nonlinearly constrained case. Instead of using a penalty function or a filter or an improvement function to deal with the presence of constraints, the search direction is determined by solving a convex quadratically constrained quadratic program to obtain good iteration points.

3: %AN%Issues arising with this procedure concerning the bundling process as well as in the line search are discussed.

4: Furthermore, global convergence of the method is shown under certain mild assumptions.

5: \end{abstract}

6: