The best way to conference proceedings by Francis Academic Press

Web of Proceedings - Francis Academic Press
Web of Proceedings - Francis Academic Press

Research on Iterative Learning Neural Network Control for Manipulator

Download as PDF

DOI: 10.25236/IIICEC.2019.057


Gailian Zhang

Corresponding Author

Gailian Zhang


with the rapid development of science and technology, robot technology has also developed rapidly and is widely used in various industries, such as industry, aerospace, military, medical and other industries. The robotic arm is a mechanical device that simulates a human arm, and is the main execution mechanism of a robot. The manipulator system itself is a complex system with nonlinear, strong coupling and interference. Moreover, in practice, due to the very complicated working conditions, it is difficult to establish an accurate mathematical model of the robotic arm system, such as uncertain load, uncertain system parameters, or even no model at all. In addition, the transmission system from the output shaft of the motor to the execution axis of the robot arm inevitably produces flexibility, so the control of the robot arm considering the flexibility of the joint has also become a hot and difficult point of research. In addition, in order to save costs or reduce measurement errors, the robotic arm in industrial applications cannot measure some state variables. Designing an observer has also become an important part of the controller design. The research in this paper first considers a sliding mode adaptive controller based on a first-order filter observer under the conditions of uncertain system parameters and some state quantities that cannot be measured. Finally, based on the completely model-free situation in the actual process, an adaptive observer of bp neural network is designed, and based on the universal approximation principle of nonlinear terms of the neural network, the inversion control method is adopted the tracking of the movement track of the robot arm is realized.


Manipulator; Iterative Learning; Neutral Network