An Adaptive Backstepping Terminal Sliding Mode Control for Stewart Platforms

This paper presents an adaptive backstepping terminal sliding mode controller for tracking control of Stewart platforms. By designing an integral nongsingular fast terminal sliding surface, the system can achieve finite-time convergence, small tracking errors, high robustness over un-modeled dynamics, and time-varying external disturbances. In addition, the backstepping control law with an adaptive gain based on the Lyapunov stablity theory guarantees system's globally asymptotic stability without precise knowledge of the upper bound of the uncertainty. For the control design, the robot's dynamic model was first established and formulated in the active joint space. The effectiveness of the controller is verified through simulation in comparison with a computed-torque controller. The simulation results show that the proposed controller has a superior performance of small tracking errors to that of a computed torque method, and it is robust to model parameter variations (up to 30%) and time-varying uncertainties.