Parallel Network-Based Sliding Mode Tracking Control for Robotic Manipulators with Uncertain Dynamics

Robot dynamics model uncertainty and unpredictable external perturbations are important factors that influence control accuracy stability. To accurately compensate for the in sliding mode (SMC), a new parallel network (PCR) is proposed this paper. The parallelizes radial basis function convolutional neural network, which gives it advantage of making full use one-dimensional data fitting results two-dimensional feature information, realizing deep learning multidimensional improving model’s compensation anti-interference ability. Meanwhile, based on integration adaptive techniques gradient descent, weight update algorithm designed to realize online PCR networks under loss-free functions. Then, controller (PCR-SMC) established. model-free intelligent robot accomplished without knowledge predetermined upper bounds. Additionally, stability analysis system proved by Lyapunov theorem. Lastly, tracking simulations performed two trajectories. demonstrate high-precision performance comparison with RBF-SMC controller.