Key points are not available for this paper at this time.
A novel tracking control system with radial basis function neural network (RBF) for the Selective Compliance Articulated Robot Arm (SCARA) robot arm with 4 degrees of freedom is presented, which integrates the incremental PID controller and sliding mode robust controller (RDRBF). The neural network allows the control system to learn the uncertain model of robot on line, so that there is no need to know the exact dynamic model of robot in advance. The combination of the sliding mode robust controller with incremental Proportional-Integral-Derivative controller (PID controller) guarantees the control system's performance of a better stability, accuracy and disturbances rejection. The contrastive experiments indicate that the proposed control system performs better for both the position tracking and speed tracking, compared to the traditional PID controller and pure neural network controller for trajectory tracking especially with an extra strong interference or dynamic model uncertainties.
Guo et al. (Thu,) studied this question.