Two-degree-of-freedom robotic arm based on RBF neural network approximation control method research

The robotic arm in industrial badlands has the problem of insufficient running accuracy or easy jitter in movement, and the anti-interference ability is not excellent enough. In order to solve this problem, this paper proposes a control method based on radial basis neural network approximation. In this paper, firstly, the kinematic, dynamics modeling of the two-degree-of-freedom robotic arm is carried out. Secondly, the control law of radial basis neural network is designed and simulation experiments are conducted for the robotic arm, and the improved joint angle and angular velocity running trajectory and the desired trajectory are observed experimentally. The results show that the trajectory tracking is excellent, and the error can be decreased and stabilized at a very small value with time. It can be concluded that the effectiveness and superiority of the algorithm in this paper for the control of the two-degree-of-freedom robotic arm.