ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 6. Vol. 26. 2020

DOI: 10.17587/it.26.354-359

O. N. Krakhmalev, Ph.D., Associate Professor, olegkr64@mail.ru, Bryansk State Technical University, Bryansk, 241036, Russian Federation

Algorithmization of Parallel Computing in a Dynamic Model of Robot Manipulation Systems

The task of developing an algorithm for parallel computing in a dynamic model of robot manipulation systems obtained by the Lagrange-Euler method is considered.
The algorithm is built on the basis of the object-oriented approach by the decomposition method. In the structure of the dynamic model highlights the fundamental components. Geometric and inertial models of manipulation systems are defined as such components. Using the language of object-oriented programming, classes are created that describe the structure of the geometric and inertial models. Independent objects can be initialized based on the created classes. By combining objects in various combinations, object schemas can be obtained. The initial dynamic model is considered as an instruction for building the corresponding object schema. Algorithmization of parallel computing is based on the independence of the calculation of data objects of classes. The use of parallel computational algorithms in modeling problems requires a sufficient number ofprocessors to implement such calculations. This requirement can be implemented on the basis of multi-chip reconfigurable computer systems created using programmable logic integrated circuits. Such computing systems can adapt to the structure of the problem to be solved and provide the implementation of complex multi-level parallel algorithms.
Keywords: robots, manipulation systems, dynamic models, mathematical modeling, object-oriented approach, computational algorithms, parallel computing

P. 354–359

To the contents