Manipulators 3D printing trajectory tracking control combined with RBFNNs and visual feedback

Abstract

Manipulators are used in the 3D printing due to their flexibility and wide working range. However, due to the tandem structure themselves, the motion accuracy of the end-effector is low, which affects the geometric accuracy and quality of printed products. Therefore, it is proposed to use visual servo control to improve the trajectory accuracy, and combine the radial basis function neural networks (RBFNNs) to approximate the robot dynamic parameters, so as to avoid the influence of the manipulators’ mass inertia and gravity properties on the motion accuracy under high speed and heavy load. Aiming at the visual measurement feedback method, a combined method of error similarity theory in continuous space and inverse distance weighting is proposed to estimate the actual error of the endeffector. The controller of RBFNNs dynamics feedforward coupled visual servo sliding mode control (SMC) is designed, and Lyapunov stability theory is used to prove the stability of the control system. Combined with Matlab/Simulink to build a simulation model to verify the effectiveness of the controller.