Projects and Grants
FR-TI1/420
Integrated software system for optimization and control of technological processes and prediction of production quality for large technological complexes with use of advanced mathematical methods, simulation and artificial intelligence.
Project is oriented on development and realization of integrated software system for optimization of control of technological processes and prediction of performance quality for large technological complexes with use of advanced mathematical methods, simulation and artificial intelligence.
Years: 2009 - 2012
Program: MPO Ministry of Industry and Trade (MPO), FR-TIP (2009 - 2017)
Company: PIKE AUTOMATION spol. s r.o.
People:
- Vaclav Snasel
- Jan Platos
- Pavel Kromer
- Tomas Novosad
- Martin Radvansky
- Tomas Novosad
GA102/09/1494
New methods of data transmission based on turbo code
Throughout the past decade, the discovery of turbo codes has been considered as one of the most successful researches whose impact significantly enhanced the quality of
modern communication systems. Turbo codes offer the best compromise between structure (concatenation) and randomness created by the interleaver. Our research and development has been focused on the area of turbo code interleaver optimization in the recent time. We have investigated the ability of genetic algorithms, perhaps the most used natureinspired optimization technique, to find a turbo code interleaver with above average performance. A new modification of genetic algorithms, designed to deal with constraints of a combinatorial optimization problem such as interleaver optimization task, was developed.
Years: 2009 - 2011
Program: GA0 Czech Science Foundation (CSF)
People:
- Vaclav Snasel
- Jan Platos
- Pavel Kromer
- Nabil Oudane
Energy Markets Splitting and Matching
Contracted Research and Development project for Indra Czech Republic.
Problem: Design and implement matching module for market splitting on energy market of Czech Republic and Slovakia.
Solution: Detailed design of the solutions for independent implementation. Multiplatform (Win, Linux, Solaris) solution.
- Hundreds of offers/demands.
- Up to 25 hours per offer.
- Up to hundreds of units per hour per offer/demand.
- Totally more than 10 5 of units to process and evaluate every hour in a day.
Algorithm:
- Short time to process (up to 5 minutes for one day).
- Deterministic algorithm for first acceptable solution.
- Consecutive Optimization using Bioinspired algorithms.
- Strong focus on the optimization criteria.
- Selfdescriptive output.
Years: 2009 - 2011
People:
- Vaclav Snasel
- Pavel Kromer
- Jan Platos