Disturbance cancellation techniques for MIMO smooth sliding mode control

Abstract

Sliding mode control is a control method, traditionally characterized by infinite frequency switching, which is very robust to unmodeled noise and uncertainty. While sliding mode control provides very robust finite time convergence to the sliding surface, the implementation of this high frequency switching variable structure control is not generally realizable. This leads to the use of continuous approximations that are realizable but do not provide finite time convergence to the sliding surface and are not robust to unmodeled noise and disturbances. In this paper, we present a multi-input, multi-output smooth sliding mode control design which is a true (not approximated) sliding mode control design with finite time convergence to the sliding surface and is more robust to unmodeled disturbances and plant dynamics than continuous approximations. However, the Smooth Sliding Mode Controller (SMC) is not as robust to these disturbances as the traditional SMC. In order to improve the robustness of the Smooth SMC, we also investigate two methods, a disturbance estimator and an integral sliding surface, to enhance the robustness of the Smooth SMC. The enhanced robustness of the Smooth SMC using both of these methods is demonstrated using a numerical example.