McCormick | Northwestern Engineering _

Andreas Wächter

Professor of Industrial Engineering and Management Sciences

Research Overview

I'm interested in the development of practical numerical algorithms for computational optimization. This includes the design and theoretical analysis of new methods, as well as their software implementation and practical application.

My research concentrates on optimization problems that involve nonlinear objective and constraint functions.

The broad themes are:

Research Projects

Interior point methods for nonconvex nonlinear constrained optimization

Starting with my PhD thesis advised by Larry Biegler, I developed Ipopt, a general purpose open-source optimization solver for large-scale nonlinear optimization. Ipopt was first released as FORTRAN code in 2000, and then reimplemend in C++ in 2005-2006 at IBM by Carl Laird and myself. Ipopt has been widely used in numerous application and has been integrated in various commercial software.

My work on Ipopt was awarded the 2009 INFORMS Computing Sociery Prize and the 2011 Wilkinson Prize for Numerical Software.

Various topics related to this research:

Nonlinear two-stage optimization problems

Optimization problems in electrical power systems

Simulation optimization

Optimization problems with nonlinear chance constraints

Optimization problems with complementarity constraints

Derivative-free optimization

Nonsmooth optimization

Sequential quadratic programming methods

Algorithms for mixed-integer nonlinear optimization

During my time at IBM, I was part of a team that developed the Bonmin and Couenne open-source solvers for nonlinear optimization