ZIB’s research on integrality constraints is based on a successful interplay between problem specific investigations, which can be generalized, and advances in solution methods for general mixed-integer non-linear optimization problems (MINLPs), which help to solve specific problems.

Tremendous speed-ups in the last decade bring new applications in engineering and economics into reach. We aim at developing innovative methods for the solution of large and complex MINLPs that combine discrete-continuous, non-linear, and stochastic model components. Several state-of-the-art implementations are available in source code.