A Primal-Dual Trust-Region Algorithm for Minimizing a Non-convex Function Subject to General Inequality and Linear Equality Constraints

A. R. Conn, N. I. M. Gould, D. Orban, and Ph. L. Toint

A new primal-dual algorithm is proposed for the minimization of non-convex objective functions subject to general inequality and linear equality constraints. The method uses a primal-dual trust-region model to ensure descent on a suitable merit function. Convergence is proved to second-order critical points from arbitrary starting points. Preliminary numerical results are presented.

Technical Report RAL-TR-1999-054 Rutherford Appleton Laboratory Chilton, Oxfordshire, England

Contact: n.gould@rl.ac.uk