Control System of the RobotManipulator with Use of Neural Network Algorithms of Restriction of Work Area of the Gripper
DOI:
https://doi.org/10.14529/ctcr170404Keywords:
robot-manipulator, artificial neural network, perceptron, training set, control systemAbstract
This article covers control system architecture of industrial robot (manipulator), designed to work in heavy nuclear fields. To increase safety of manipulator control and moving an additional level of monitoring gripper position has been added to control system. This level includes artificial neural network, based on perceptron with output signal in range [0;1], which is used as a coefficient of transferring manual controls from joysticks to internal loops of control system. Outlined the way of preparing teaching data set for neural network and results of control system math modeling.
References
Юревич, Е.И. Основы робототехники / Е.И. Юревич – СПб.: БХВ-Петербург, 2005. – 416 с.
Назаров, А.В. Нейросетевые алгоритмы прогнозированиия и оптимизации систем / А.В. Назаров, А.И. Лоскутов. – СПб.: Наука и техника, 2003. – 384 с.
Application of Neural Networks and Other Learning Technologies in Process Engineering / I.M. Mujtaba, M.A. Hussain (eds.). – Imperial College Press, 2001. – 405 p.
