Universita' di Firenze
Dipartimento di sistemi e informatica
Via di S.Marta 3, 50139 Firenze
Tel. ++39 055 4796 464
This is the home page of the operational research group within the Faculty of Engineering of the University of Florence.
This laboratory is dedicated to the memory of Gerardo Poggiali
Professors and PhD Students:
Ing. Andrea Cassioli(PhD)
Ing. David Di Lorenzo (PhD student)
Ing. Mirko Maischberger (PhD student)
External collaborators:
Ing. Paolo Frosini(PhD)
July 22, 2009: METSlib Metaheuristic library Mirko Maischberger from GOL uploaded the METSlib project to COIN-OR. METSlib is a metaheuristics modeling framework and optimization toolkit in C++. Currently, the toolkit implements the basics of some metaheuristic algorithms: Local Search, Simulated Annealing, and Tabu Search. Users can create their own models and use the built-in algorithms to find solutions. Also, users can implement other metaheuristics by using existing modules as building blocks.
A course on crew rostering and staffing for airlines has been held in Florence, May 05 - May 07, 2008 CIME Summer School on Nonlinear Optimization -Grand Hotel Villa San Michele (Cetraro,Italy), July 1-7, 2007. See also our CIME school web page We started a Monthly Seminar series (two speakers in the same day) as well as an internal series of seminars (one every two tuesdays). December 2006: New book anouncement: Modelli di Ottimizzazione per le Decisioni June 22, 2006: Many new putative optima have been discovered for Binary Lennard-Jones atomic clusters; they are going to be published on the Cambridge Cluster Database Jan 14, 2006: We won the Circle Packing Contest - an optimization contest with 155 participants and 27 490 total submissions! 30.06.2005: Further new putative optima have been discovered for the problem of optimally packing circles in the unit square. We improved putative gloablly optima configuration for N=53, 59, 66, 68, 73, 77, 78, 85, 86, 88, 100, 101, 106, 108, 113, 115, 116, 130, 133, 134, 135, 146, 155, 157. See the official repository of putative optimal packings
17.2.2003: 5 new putative global optima have been found for Morse Clusters at rho=8 using our two-phase monotonic basin hopping algorithm. The new optima are those for 24, 25, 45, 48 and 51 atoms See the report: New global optima for Morse clusters at $\rho=8$ |
Past news are available.
Stochastic global optimization: theory
Large Scale Global optimization algorithms
Optimization in protein-protein docking
Protein folding
Distributed optimization
Applied operations research: transportation, health care, simulation
Mini HowTo of the laboratory
GOL internal database of scientific literature
GOL svn server
1 SUN Workstation running Linux
5 PC's running Linux (Debian)
3 PC's running Win98
2 Laptop PC's (Win98/XP)
1 Linux Cluster (one master + 4 slave machines)
Main software licences available:
Knitro professional edition, 1 floating licence
AMPL - professional edition (Linux/Win)
Cplex - professional edition (Linux)
MS Office prfessional (Win)
MS Visual Studio (Win)
Qualcomm Eudora Pro (Win)
Adobe acrobat (Win)
Arena simulation software (20 network licences) (Win)
Extend simulation software student edition (Win)
S-Plus 2000 statistical software (Win)
EMME/2 traffic network simulation and analysis (Win)
LEDA - library for efficient data types and algorithms (Linux)
OPT++ - a C++ library for nonlinear optimization
SNOPT - large scale sparse nonlinear optimization code
ESOPO - an environment for Large Scale Optimization
Software developed in the laboratory
GOL (Global Optimization Library): a C++ library of classes for large scale global optimization. It contains data structures for atoms, water molecules, circles-in-a-square as well as general functions
L-BFGS: a C++ port of an excellent quasi-newton method
ljopt: global optimization of Lennard-Jones molecular clusters
morseopt: global optimization of Morse atomic clusters
dock: a stochastic global optimization method for rigid protein-protein docking
Test Functions - revised version: a C++ class for generating test functions for global optimization.