A Framework for Modeling and Solving Mixed-Integer Nonlinear Programming Problems on Networks
Lamatto++ is a framework to model mixed-integer nonlinear programming problems especially on (but not limited to) networks. Lamatto++ is completely written in C++ and has interfaces to CPLEX, GUROBI, SCIP and GAMS, where the latter provides access to a large collection of solvers for roughly all kinds of mathematical programming problems. Lamatto++ major features are a unified modeling API and a single interface to all the above mentioned solvers. Lamatto++ comes with extendable data structures to handle graphs and networks and provides automatic unit conversion for physical values. For data I/O and data exchange with third party software, Lamatto++ makes use of the XML parsing library Xerces.
More details on the aim, functionality, and software design of the Lamatto++ software framework can be found in the references listed below.
We gratefully thank all members of the ForNe project for their help in improving Lamatto++.
Special thanks go to Marc Pfetsch for implementing the Lamatto++ SCIP interface and to Christoph Conrads, Armin Fügenschuh, Ralf Gollmer, Christine Hayn, Benjamin Hiller, Jesco Humpola, Imke Joormann, Thomas Lehmann, Ralf Lenz, Debora Mahlke, Andris Möller, Djamal Oucherif, Henrik Pilz, Lars Schewe, Martin Schmidt, Robert Schwarz,Jonas Schweiger, Claudia Stangl, Jacint Szabo, Stefan Vigerske, Bernhard Willert, and Andrea Zelmer for their active participation.
State of development
- Technical Transient Optimization of Gas Networks
- Odysseus - Diskret-kontinuierliche Optimierung komplexer dynamischer Wasserver- und entsorgungssysteme
Since 2009 Lamatto++ was heavily enhanced when it became a substantial part of the project:
Up to now, Lamatto++ is still under development and not publicly available.
- A. Morsi.
Chapter 5 of Solving MINLPs on Loosely-Coupled Networks with Applications in Water and Gas Network Optimization.
Ph.D. thesis, University of Erlangen-Nuremberg, 2013.