Neural sliding mode control for turntable servo system with unknown deadzone

This paper proposes a neural sliding mode control scheme for a turntable servo system with an unknown nonlinear dead-zone input. Based on the differential mean value theorem, the nonlinear dead-zone is represented as a simple time-varying system and the dead-zone inverse compensaiton approach is avoided. All the unkown nonliearities of the servo system are approximated by using a simple sigmoid neural network, and the approximation errors are compensated by adding robustness terms in the controller structure. The proposed controller can guarantee the fast tracking performance of the turntable servo system in the presence of system uncertainties and unknown dead-zone. Finally, comparative experiment with PID controller is conducted to show the effectiveness and superior performance of the proposed method.