A new super-twisting algorithm with state extension of sliding variable toward fast convergence of high accuracy

In view of the hard issue that high steady-state error of modified super-twisting algorithm under the condition of fast convergence, a scheme of extended state Super-twisting sliding mode control is presented in this paper so as to achieve a fast convergence of high accuracy. At first, an extended state model of Super-twisting algorithm (STA) is built, in which the extended state variable can enhance the steering and regulation capacity of STA, thus the steady-state error of sliding variable can be driven down effectively and even vanished. In this way, the convergence accuracy and robustness can be strengthened greatly. Then the extended state Super-twisting algorithm (ESSTA) is proved to be asymptotically stable by Lyapunov method, and its performance is validated by simulation experiment of position tracking on DC servo system.