Wyniki wyszukiwania - Biblioteka Nauki

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

Znaleziono wyników: 3

Liczba wyników na stronie

Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Modeling and analysis of dynamic properties of small electromechanical drive systems

100%

Kratochvil C. , Brezina T.

2005

tom z. 28

89-98

Analysis of dynamic properties of interactive drive systems is lately the object of interest among researches from both academic and industry fields. This interest is reflection of mutually contradicting requirements put into the development of current drive system, mainly the drives itself. On one hand we see the performance and failure-free operation requirements, on the other hand the reduction of energy input, size minimization and operation automation. The paper introduces some of the problems involved in this field.

Predicted congestions never occur. On the gap between transport modelling and human behaviour

80%

Frey H. , Leth U. , Mayerthaler A. , Brezina T.

tom T. 6, z. 1

73-85

This paper presents an introduction to meso-scale transport modeling and issues of human behaviour in transport systems. Along with other examples of the human ability to learn in transport systems we look at the comparison of real life data and the prediction of modeling tools for the closure of Vienna’s inner ring road during the 2008 European Football Championship (EURO 2008). Some light is shed on the scientific question, whether currently used modeling tools are able to adequately reproduce the real-life behaviour of human beings in the transport system and should be used for transport policy decision making.

Self-learning controller of active magnetic bearing based on CARLA method

80%

Brezina T. , Turek M. , Pulchart J.

tom R. 53, nr 1

6-9

The active magnetic bearing control through analytically designed linear PD regulator, with parallel nonlinear compensation represented by automatic approximator is described in this contribution. Coefficient (parameter) values come from actions of Continuous Action Reinforcement Learning Automata (CARLAs). Influence of CARLAs parameters to learning is discussed. Parameters influence is proved by simulation study. It is shown that learning improvement can be reached by selecting appropriate parameters of learning.

W artykule przedstawiono sterowanie aktywnego łożyska magnetycznego za pomocą analitycznie dobranego regulatora PD z nieliniową kompensacją równoległą. Współczynniki kompensacji są wyznaczane automatycznie z użyciem metody CARLA (Continuous Action Reinforcement Automata). Zbadano wpływ parametrów metody na proces uczenia się kompensatora w oparciu o eksperymenty symulacyjne. Wykazano, że właściwy dobór parametrów metody prowadzi do poprawienia skuteczności procesu uczenia się.