A New Injection for Disturbance Estimation Subject to Measurement Noise

This paper investigates a novel extended state observer incorporating static dead-zone injection, aiming to mitigate the impact of high-frequency noise on the system by introducing zero dynamics near y-\hat{x}_{1}. In this work, we extend the application of dead-zone injection to disturbance estimat...

Full description

Saved in:
Bibliographic Details
Published in2024 43rd Chinese Control Conference (CCC) pp. 775 - 779
Main Authors Hao, Liwei, Zhang, Yingchun
Format Conference Proceeding
LanguageEnglish
Published Technical Committee on Control Theory, Chinese Association of Automation 28.07.2024
Subjects
Dead zone
Disturbance estimation
Estimation
Extended state observer
High-gain method
Measurement noise
Noise
Observers
Sensitivity
Simulation
Uncertainty
Upper bound
Online AccessGet full text

Cover

More Information
Summary:This paper investigates a novel extended state observer incorporating static dead-zone injection, aiming to mitigate the impact of high-frequency noise on the system by introducing zero dynamics near y-\hat{x}_{1}. In this work, we extend the application of dead-zone injection to disturbance estimation and propose a new parameter design approach for high-gain extended observers. This method focuses on analyzing the characteristics of the high-gain extended state observer, and designing a new dead zone using the high-gain parameters, and provide proof of its ultimate upper bound considering bounded measurement noise. The observer proposed in this paper exhibits superior noise resistance compared to the standard high-gain observer, while avoiding the introduction of excessively intricate new parameters. Based on the establish conclusions regarding the practical convergence of the system, we present a numerical example to illustrate the advantages offered by this design. Simulation results comparing our proposed observer with conventional extended state observers are provided under measurement noise conditions, demonstrating effective suppression of measurement noise through our new injection.
ISSN:1934-1768
DOI:10.23919/CCC63176.2024.10662832