An investigation of various controller designs for multi-link robotic system (Robogymnast)

Abdul samad, Bdereddin ORCID: https://orcid.org/0000-0002-9140-5024, Mohamed, Mahmoud ORCID: https://orcid.org/0000-0001-9386-7495, Anayi, Fatih ORCID: https://orcid.org/0000-0001-8408-7673 and Melikhov, Yevgen ORCID: https://orcid.org/0000-0002-9787-5238 2022. An investigation of various controller designs for multi-link robotic system (Robogymnast). Knowledge 2 (3) , pp. 465-486. 10.3390/knowledge2030028

PDF - Published Version Available under License Creative Commons Attribution. Download (4MB)

Abstract

An approach to controlling the three-link Robogymnast robotic gymnast and assessing stability is proposed and examined. In the study, a conventionally configured linear quadratic regulator is applied and compared with a fuzzy logic linear quadratic regulator hybrid approach for stabilising the Robogymnast. The Robogymnast is designed to replicate the movement of a human as they hang with both hands holding the high bar and then work to wing up into a handstand, still gripping the bar. The system, therefore has a securely attached link between the hand element and the ‘high bar’, which is mounted on ball bearings and can rotate freely. Moreover, in the study, a mathematical model for the system is linearised, investigating the means of determining the state space in the system by applying Lagrange’s equation. The fuzzy logic linear quadratic regulator controller is used to identify how far the system responses stabilise when it is implemented. This paper investigates factors affecting the control of swing-up in the underactuated three-link Robogymnast. Moreover, a system simulation using MATLAB Simulink is conducted to show the impact of factors including overshoot, rising, and settling time. The principal objective of the study lies in investigating how a linear quadratic regulator or fuzzy logic controller with a linear quadratic regulator (FLQR) can be applied to the Robogymnast, and to assess system behaviour under five scenarios, namely the original value, this value plus or minus ±25%, and plus or minus ±50%. In order to further assess the performance of the controllers used, a comparison is made between the outcomes found here and findings in the recent literature with fuzzy linear quadratic regulator controllers.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Engineering
Additional Information:	License information from Publisher: LICENSE 1: URL: https://creativecommons.org/licenses/by/4.0/, Type: open-access
Publisher:	MDPI
Date of First Compliant Deposit:	10 October 2022
Date of Acceptance:	26 August 2022
Last Modified:	07 Jul 2023 16:53
URI:	https://orca.cardiff.ac.uk/id/eprint/153278

Actions (repository staff only)

Edit Item