Project Details
Výzkum metod modelování a simulace inteligentních systémů
Project Period: 1. 1. 2006 – 31. 12. 2008
Project Type: grant
Code: GP102/06/P309
Agency: Czech Science Foundation
Program: Postdoktorandské granty
intelligent systems, algorithmic game theory
The project interconnects computer modeling, simulation, game theory and
applications of strategic decision models in economic problems. The project deals
with a general methodology of modeling the decision situations, which was later
extended to multi-dimensional decision making. Regarding the multi-dimensional
decisions, three basic forms of the decision combinations were shown together
with a methodology of their composition. Some of the studied problems are
directly connected to the Ph.D. thesis of the author. That is a methodology of
game modeling based on a special application of the Automatic Information Net
(AIN). A case-study published in one of the papers clearly presented that a game
model implemented in the AIN manner can significantly speed up the computation
when used in repeated experimenting (the model is treated as persistent and it
evaluates just the consequences of the modified inputs). The algorithmic
approaches to solving of extremely large multi-player games (e.g., the state
space contains 10^30 members) can be without any doubt regarded as the main
contribution of the project. Two methods (CE-Solver, FDDS) have been developed in
order to allow an efficient reduction of state space of multi-player games and to
determine the correlated equilibrium (Aumann, 1974). Both methods are
parallelizable. The main theoretical outcomes of the project were summarized in
form of two large scientific papers intended for peer-reviewed journals (one of
them is currently in the review process). All the theoretical results have been
verified during the practical implementation of models of strategic behavior of
producers and consumers in the area of electro-energy commodity markets in Czech
republic and Central Europe in cooperation with EGU Brno Ltd. Simulation results
made by these models are required by the government organizations responsible for
the electricity network management and maintenance (OTE, CEPS, MPO, ERÚ). The
grant results thus serves for the management of the national electricity
transmission system.
2008
- HRUBÝ, M. Algorithmic Approaches to Game-theoretical Modeling and Simulation. AUCO Czech Economic Review, 2008, vol. 2, no. 3,
p. 268-300. ISSN: 1802-4696. Detail - HRUBÝ, M.; ČAMBALA, P. Efficient Computing of Correlated Equilibria in Multi-Player Games. Proceedings of 12th IASTED Conference on Artificial Intelligence and Soft Computing. Calgery: ACTA Press, 2008.
p. 185-191. ISBN: 978-0-88986-755-0. Detail
2007
- HRUBÝ, M. Modelling the Structured and Complex Decision Situations. Proceedings of the 6th EUROSIM Congress on Modelling and Simulation. Ljubljana: 2007.
p. 236-245. ISBN: 978-3-901608-32-2. Detail - HRUBÝ, M.; BEDNÁŘ, J. Automated Optimization of Models Based on Game Theory. Proceedings of International Symposium on Innovations in Intelligent Systems and Applications. Istanbul: IEEE Turkey, 2007.
p. 134-138. Detail
2006
- HRUBÝ, M. Using the HELEF Language to Model the Strategic Games. Proceedings of XXVIIIth International Autumn Colloquium ASIS 2006: Advanced Simulation of Systems. Ostrava: 2006.
p. 121-126. ISBN: 80-86840-26-3. Detail - HRUBÝ, M. Výuka GIS na FIT VUT v Brně. Proceedings of Geoinformatics CTU 2006. Proceedings of the Workshop Geoinformatics FCE CTU 2006. Praha: Vydavatelství ČVUT, 2006.
s. 10-15. ISSN: 1802-2669. Detail - HRUBÝ, M.; ČAMBALA, P.; TOUFAR, J. Modelování spolupráce elektrizačních soustav v evropském regionu. Energetika, 2006, roč. 2006, č. 11,
s. 343-346. ISSN: 0375-8842. Detail - HRUBÝ, M.; TOUFAR, J. Modelling the Electricity Markets using Mathematical Game Theory. Proceedings of the 15th IASTED International Conference on APPLIED SIMULATION AND MODELLING. Proceedings of the 15th IASTED International Conference on Applied Simulation and Modelling. Calgery: ACTA Press, 2006.
p. 352-357. ISSN: 1021-8181. Detail