MODELING OF AN ADAPTIVE ARTIFICIAL NEURALNETWORK BASED POSITIONING SYSTEM FOR A TWO-LINK KINEMATIC CIRCUITS
Abstract
The application of a pretrained artificial neural network as a regulator for positioning of free links of a two-link kinematic chain formed by series- connected spherical kinematic pairs of the third class is modeled. The issues of mathematical modeling of the mechanical component of the system, including the mechanism for changing the position of the links; formation of a training sample of coordinates of the position of the final free link; choosing the structure of an artificial neural network and training it to solve the inverse problem of kinematics; analysis of the accuracy of the resulting positioning system, are considered.
About the Authors
S. K. Shulgin
Luhansk State University named after Vladimir Dal
Russian Federation
Russian Federation
Shulgin S.K. - candidate of technical sciences, associate professor of the department "Information and control systems"
D. O. Sinepolsky
Luhansk State University named after Vladimir Dal
Russian Federation
Russian Federation
Sinepolsky D.O. - senior lecturer of the Department "Information and Control Systems"
A. V. Balalayechnikov
Luhansk State University named after Vladimir Dal
Russian Federation
Russian Federation
Balalayechnikov A.V. - senior lecturer of the Department "Information and Control Systems"
References
1. "Robotics Technology and Flexible Automation 2nd Edition" S.R. Deb. McGraw Hill. New Delhi. 2010 ISBN 0-07-007791-6
2. "Robot Modeling and Kinematics" Manseur, Rachid. DaVinci Engineering Press. Boston, MA. 2006 ISBN 1-58450-851-5 Ch.4-5
3. P. Moubarak, et al., Modular and Reconfigurable Mobile Robotics, Journal of Robotics and Autonomous Systems, 60 (12) (2012) 1648—1663.
4. Шульгин С.К. Моделирование адаптивной системы позиционирования звена сферической кинематической пары на основе искусственной нейронной сети [Тект] / Шульгин С.К., Синепольский Д.О. // Вестник ЛГУ им. В. Даля / ГОУ ВО ЛНР ЛГУ им. В. Даля. – Луганск, 2022. – №6.
5. Понарин Я. П. Элементарная геометрия. В 2 т. — М.: МЦНМО, 2004.— ISBN 5-94057-170-0.
6. S. Tejomurtula and S. Kak : «Inverse kinematics in robotics using neural networks» Information Sciences, Vol. 116, pp. 147–164, 1999.
7. Y. Maeda, T. Fujiwara and H. Ito : «Robot control using high dimensional neural networks» Procs. of SICE Annual Conference 2014, pp. 738–743, 2014.
8. Practical optimization, Philip E. Gill, Walter Murray and Margret H. Wright, Academic Press Inc. (London) Limited, 1981.
Review
For citations:
Shulgin S.K., Sinepolsky D.O., Balalayechnikov A.V. MODELING OF AN ADAPTIVE ARTIFICIAL NEURALNETWORK BASED POSITIONING SYSTEM FOR A TWO-LINK KINEMATIC CIRCUITS. Social-economic and technical systems: research, design and optimization. 2023;(2):122-130. (In Russ.)
Views:
3
Email this article 