Accelerated Adaptive Second Order Super-Twisting Sliding Mode Observer

Abstract

An accelerated second-order super-twisting sliding mode observer with an adaptive gain is proposed for a typical nonlinear system. The key contribution of this algorithm is that the rate of convergence of observation error is accelerated remarkably by introducing “system damping.” Chattering issue is attenuated with a satisfactory performance compared with conventional sliding mode observer. Furthermore, adaptive gain can vary with deviation between the trajectory and the sliding mode switching manifold dynamically so that overshoot can be reduced. The novel observer is proven mathematically to be convergent in a finite time. Finally, an example of nonlinear system is given to verify the performance.