Generalized Super-Twisting Observer for a class of interconnected nonlinear systems with uncertainties

TL;DR

This paper introduces a generalized super-twisting observer for interconnected nonlinear systems with uncertainties, demonstrating finite-time convergence and improved estimation accuracy over traditional methods.

Contribution

The paper extends the super-twisting observer to a broader class of interconnected nonlinear systems with uncertainties, using a nonsmooth Lyapunov function for convergence proof.

Findings

GSTO achieves finite-time convergence despite uncertainties

Discontinuous term is crucial for exact estimation

Outperforms High Gain observer in case study

Abstract

The Generalized Super-Twisting Observer (GSTO) is extended for a strongly observable class of nonlinearly interconnected systems with bounded uncertainties/perturbations. A nonsmooth strong Lyapunov function is used to prove the finite-time convergence of the proposed observer to the true system's trajectories, in the presence of the uncertainties. A case study on the interaction between two food production systems is presented, comparing the proposed observer with the High Gain observer. The results emphasize the critical role of the GSTO's discontinuous term in achieving exact estimation.