Hybrid control of robotic manipulator by neural network model. 3rd rtport. Control of 2 D.O.F. manipulators by neural network using fuzzy set theory.
Toshio Fukuda, Takanori Shibata, Masatoshi Tokita, Toyokazu Mitsuoka · TRANSACTIONS OF THE JAPAN SOCIETY OF MECHANICAL ENGINEERS Series C · 1991
This paper deals with the neural network application for the robotic motion control where the controller is applicable to the position and force control of robotic manipulators. The proposed "Neural Servo Controller" is based on the neural network which has integrated time delay elements, so that the neural network can learn the dynamical system. Moreover, we proposed the "Fuzzy Turbo", which employs the fuzzy set theory to avoid atagnation and to have insensitive characteristics at a stable extreme on learning so that the neural network can learn the dynamical system quickly. Simulations are carried out for the case of the force control of a two-dimensional robotic manipulator which handles unknown objects and for the case of its trajectory control which handles unknown pay-loads. The results show that the proposed "Neural Servo Controller" is applicable to the nonlinear and dynamical system and that the proposed "Fuzzy Turbo" avoids stagnation in learning, so the neural network can learn guiculy.