A new perturbation estimator in a sliding mode control system with application to robot control

In this work, a new perturbation estimator associated with a sliding mode controller is proposed to enhance control performance of linear or non-linear systems under perturbations such as parameter uncertainties and extraneous disturbances. The estimator is designed by adopting an integrated average value of the imposed perturbation over a certain sampling period. Subsequently, in order to improve control performance, a so-called perturbation filtering condition (PFC) is introduced for the estimator. By applying the PFC in which useful components of the imposed perturbations are kept, control performance can be substantially improved without additional control input. Some benefits of the proposed methodology are demonstrated on a two-link planar robotic manipulator. The position tracking control performances of the manipulator are evaluated in the time domain and a comparative work between the proposed methodology and conventional scheme for the perturbation estimation is undertaken.