Design of Remote Control Algorithm for Robot Arm Based on RBF Neural Network

• Published in: 2023 2nd International Conference on Artificial Intelligence and Autonomous Robot Systems (AIARS) pp. 354 - 358
• Main Authors: Ding, Juan, Huang, Yuejuan, Huo, Chunyan, Li, Tianshu, Zeng, Fanju
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2023
• Subjects: Arms; Delays; Kinematics; Machining; Manipulators; Mechanical arm; RBF neural network; Real-time systems; Remote control algorithm; Training
• DOI: 10.1109/AIARS59518.2023.00078

The robot arm is a complex structure with multiple degrees of freedom. The forward and inverse kinematics of the robot arm must be calculated when the robot arm is remotely controlled. However, due to the special nature of the environment in which space robotic arms are remotely controlled, such as free floating and weightlessness, the difficulty of designing space robotic arm remote control has greatly increased. It is difficult to directly promote the remote control method of ground fixed base robotic arms. At the same time, how to apply remote control theory and methods to minimize the energy consumption of the driving system to protect the environment has become one of the key issues in research. In this regard, this article designs a remote control algorithm for the robotic arm based on RBFNN (RBF Neural Network), and establishes a mechanical borrowing remote control system in the article to adjust the operating speed of the robotic arm in real-time. Finally, simulation examples were used to verify that the method proposed in this paper has relatively high accuracy, stable network, good prediction performance, and good universality after network training. It can be seen that the adaptive control scheme based on RBF neural network adopted in this article can effectively remotely control the joints of the robotic arm to accurately complete the desired motion.