Research team of Tomas Brandejsky

Research team deals with application of novel method in the area of big data processing and analysis in the area of complex (e.g. transportation) systems. To evaluation of processing characteristics of these systems, the methods and technologies of Big data, Parallel systems, Artificial Intelligence and Computer Simulation are applied:

  • Design of computer cluster architecture for HPC, Big data and Data Analytic,
  • Implementation of parallel algorithms for OpenMP, MPI, CUDA, openACC, Chapel, Spark and Scala environments
  • Detection and analysis of object movement in the space and time, temporal data processing, distributed and NoSQL databases,
  • Information systems for transport, Data Analytic. 

Key words

Clusters – Databases – Parallel HW and SW – Artificial Intelligence – Data Analytic – Big Data – Real-time

Research Team members

  • doc. Dr. Ing. Tomas Brandejsky 
  • prof. Ing. Antonin Kavicka, Ph.D. 
  • doc. Ing. Michael Bazant, Ph.D.
  • Ing. Monika Borkovcova, Ph.D.
  • Ing. Roman Divis
  • Ing. Jan Merta 

Projects being solved 

2018 – 2022

Cooperation in Applied Research between the University of Pardubice and companies, in the Field of Positioning, Detection and Simulation Technology for Transport Systems (PosiTrans). MEYS CR, programme OPVVV

2019 – 2020

Software simulation support for determining of raiway station infrastructure capacity. TACR, programme Zeta

Selected publications

VYCITAL, J. - BAZANT, M. Train overtaking at railway stations within simulationmodels of railway lines. In 32nd European Modeling and Simulation Symposium, EMSS 2020. Janov: Dime University of Genoa, 2020. p.p. 28-34. ISBN 978-88-85741-44-7.

BULICEK, J. - BAZANT, M. Selection of railway line segments that allow occupation by more trains based on simulation. In 32nd European Modeling and Simulation Symposium, EMSS 2020. Janov: Dime University of Genoa, 2020. p.p. 242-247. ISBN 978-88-85741-44-7. 

BAZANT, M. - BULICEK, J. Assessment of partial double-tracked railway lines with focus on capacity. In 24th International Scientific Conference Transport Means 2020. Kaunas: Kaunas University, 2020. p.p. 283-288. ISSN 1822-296X (print) ISSN 2351-7034 (on-line).

BAZANT, M. - BULICEK, J. Assessment of partial double-tracked railway lines with focus on capacity. In 24th International Scientific Conference Transport Means 2020. Kaunas: Kaunas University, 2020. p.p. 283-288. ISSN 1822-296X (print) ISSN 2351-7034 (on-line).

BAZANT, M. - BULICEK, J. Assessment of partial double-tracked railway lines with focus on capacity. In 24th International Scientific Conference Transport Means 2020. Kaunas: Kaunas University, 2020. p.p. 283-288. ISSN 1822-296X (print) ISSN 2351-7034 (on-line).

BRANDEJSKY, T. Preconditions of GPA-ES Algorithm Application to Big Data. In Advances in Intelligent Systems and Computing : Artificial Intelligence and Evolutionary Computations in Engineering Systems. Vol.1056. Singapur: Springer, 2020, p.p. 485-492. ISBN 978-981-15-0198-2. ISSN 2194-5357.

KAVIČKA, A. - DIVIŠ, R. - VESELÝ, P. Railway station capacity assessment utilizing simulation-based techniques and the UIC406 method. In 32nd European Modeling and Simulation Symposium, EMSS 2020. Janov: Dime University of Genoa, 2020. s. 41-49 s. ISBN 978-88-85741-44-7.

BRANDEJSKY, Tomas. Dependency of GPA-ES Algorithm Efficiency on ES Parameters Optimization Strength. Journal of Advanced Engineering and Computation. 2019, 3(1), 304-311. DOI: 10.25073/jaec.201931.226. ISSN 2588-123X.

BRANDEJSKY, Tomas. GPA-ES Algorithm Modification for Large Data. In: Intelligent Systems Applications in Software Engineering. Cham: Springer International Publishing, 2019, 2019-09-20, p.p. 98-106. Advances in Intelligent Systems and Computing. DOI: 10.1007/978-3-030-30329-7_9. ISBN 978-3-030-30328-0.

BAZANT, M. - KAVICKA, A. - DIVIS, R. - VARGA, M. Simulation-based rail traffic optimisations applying multicriterial evaluations of variants. In Mendel. Brno: VUT, 2019. s. 139-146 s.

NOVOTNY, R. - KAVICKA, A. Hybrid simulation model supporting efficient computations within rail traffic simulations. In 31st European Modeling and Simulation Symposium, EMSS 2019. Janov: Dime University of Genoa, 2019. p.p. 16-22. ISBN 978-88-85741-26-3.

DIVIS, R. - KAVICKA, A. Computational optimizations of nested simulations utilized for decision-making support. In 31st European Modeling and Simulation Symposium, EMSS 2019. Janov: Dime University of Genoa, 2019. p.p. 80-89. ISBN 978-88-5741-26-3.

MERTA, Jan and Tomas BRANDEJSKY. Lifetime Adaptation in Genetic Programming for the Symbolic Regression. In: Computational Statistics and Mathematical Modelling Methods in Intelligent Systems. Cham: Springer International Publishing, 2019, 2019-09-20, p.p. 339-346. Advances in Intelligent Systems and Computing. DOI: 10.1007/978-3-030-31362-3_33. ISBN 978-3-030-31361-6. Also accessible from: http://link.springer.com/10.1007/978-3-030-31362-3_33

Research activities in detail

1. Design of computer cluster architecture for HPC, Big data and Data Analytics

The ideal structure of a computing system is related to its specific use and concluding demands and requirements. There are even different even requirements to e.g. computational systems for scientific computation and simulation depending on the communication intensity of the model component. Even more different are the requirements to the processing of large data or even data of Big data category size. Our team is also focussed to the building of such large data analytic tools (Data Analytic) including on-line analytic, which means data analytic in nearly real-time. Examples of such systems are systems of on-line vehicle tracking, e.g. corporate cars, monitoring of their technical state and evaluating of not only of traces and times but also detection of critical places in the infrastructure influencing their reliability.

 

2. Implementation of parallel algorithms for OpenMP, MPI, CUDA, openACC, Chapel, Spark and Scala environments

It is not possible to process data in an acceptable time and it is not possible to use parallel computer systems, especially clusters, without parallel algorithms implemented in the related programming languages and software environments. The team focuses on to design of simulation, computing and analytic algorithms optimized to parallel HW and SW including advanced algorithms of artificial intelligence, as hierarchic and hybrid evolutionary algorithms.

3. Detection and analysis of object movement in the space and time, temporal data processing, distributed and NoSQL databases

There were developed algorithms for the detection of critical points in the infrastructure for databases containing many millions and billions of records about vehicle movements.

4. Information systems for transport, Data Analytic 

Also, there was developed an evolutionary algorithm for train delay distribution function formation from large databases, which is capable to find the analytic description of related distribution functions for given train delay data set.

false