An Interior Point Algorithm for Large Scale Nonlinear Programming

Richard Byrd, Mary Beth Hribar, and Jorge Nocedal

We describe a new algorithm for solving large nonlinear programming problems. It incorporates within the interior point method two powerful tools for solving nonlinear problems: sequential quadratic programming (SQP) and trust region techniques. SQP ideas are used to efficiently handle nonlinearities in the constraints. Trust region strategies allow the algorithm to treat convex and non-convex problems uniformly, permit the direct use of second derivative information and provide a safeguard in the presence of nearly dependent constraint gradients. Both primal and primal-dual versions of the algorithm are developed, and their performance is compared with that of LANCELOT on a set of large and difficult nonlinear problems.


Report OTC 97/05 Optimization Technology Center, July, 1997.