An adaptive super-twisting algorithm based on conditioning technique

Dakai Liu, Sven Esche, Mingang Wang

Research output: Contribution to journalArticlepeer-review

Abstract

This paper presents an adaptive conditioning-technique-based super-twisting algorithm aiming at improving the convergence speed and reducing the overshoot at the same time. Compared with a recently proposed method called new modified super-twisting algorithm, in which a linear acceleration factor and a damping factor are added to achieve this goal, the proposed method has several advantages. First, the proposed method enhances the convergence performance of the system by resorting to the characteristics of the conditioned super-twisting algorithm and the adaptive gains, without changing the basic structure of the classical super-twisting controller. Thus, stability proof of this method is much simpler and more concise. Furthermore, unlike the new modified super-twisting algorithm, in which an unnatural assumption on the Lipschitz disturbance is made for the stability proof, this method can counteract not only typical bounded Lipschitz disturbances but also square-root growth disturbances. Also, a set of less conservative control gains can be obtained with the proposed algorithm than with the compared algorithm. Apart from these benefits, several simulation results illustrate that the performance of the proposed method is even better in convergence and recovering from disturbance.

Original languageEnglish
Pages (from-to)497-505
Number of pages9
JournalTransactions of the Institute of Measurement and Control
Volume44
Issue number2
DOIs
StatePublished - Jan 2022

ASJC Scopus subject areas

  • Instrumentation

Keywords

  • Super-twisting algorithm
  • adaptive control
  • conditioning algorithm
  • sliding mode control

Fingerprint

Dive into the research topics of 'An adaptive super-twisting algorithm based on conditioning technique'. Together they form a unique fingerprint.

Cite this