Optimization of non-linear robust controller for robotic manipulator system

Mohamed, Mahmoud ORCID: https://orcid.org/0000-0001-9386-7495 2023. Optimization of non-linear robust controller for robotic manipulator system. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF - Accepted Post-Print Version Download (7MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (304kB)

Abstract

This study focuses on the integration of artificial intelligence and knowledge-based systems to enhance the control of a complex multi-link mechanism. The research involves the development and examination of the Robogymnast as a platform for investigating the complexities and difficulties associated with a three-link robot system. By utilising modelling, simulation, and advanced control methods, the objective of the research is to improve the overall performance and manoeuvring capabilities of mechanisms with limited actuation, thereby contributing to the progress of robotics. A mathematical model of the acrobot movement is constructed using the Lagrange equations, representing the motion of the robotic gymnast. This presents a control challenge due to its nonlinear and multivariable nature. To address this, a discrete-time linear model is proposed that specifically concentrates on the swinging action of the Robogymnast. The system's mathematical model is linearised for the investigation, providing a way to explore the determination of state space within the system. This work proposes and examines an approach to control the triple-link Robogymnast and assess its stability. In this study, a Proportional Integral Derivative (PID) controller is implemented and compared with a Linear Quadratic Regulator (LQR) to evaluate and investigate the Robogymnast system. The study also explores factors influencing the control of swing in the underactuated three-link Robogymnast. This research also endeavours to enhance the performance of a proposed PID controller by employing two distinct algorithms, the Ant Colony Optimisation (ACO) and the Gravitational Search Algorithm (GSA), to stabilize the triple-link Robogymnast robotic system. These algorithms are utilized to fine-tune the PID controller parameters before its integration with the robot for subsequent stability response evaluation. The primary focus of the study lies in examining the application of a PID controller within a three-link robotic system. The findings indicate that the ACO algorithm with PID succeeds in enhancing the system's performance when contrasted with the GSA using a PID controller. The optimised results of the system demonstrate a significant reduction in overshoot by 95.46%, from 6.386 p.u to 0.290 p.u for the first joint. The values remain at 1 p.u for joints 2 and undergo a minor change of 3%, going from 0.309 p.u to 0.300 p.u for the third joint. The rise and settling times are also significantly reduced. Importantly, the Integral Time Absolute Error (ITAE) for the first joint is reduced by 87.63%, decreasing from 94.180 to 11.650. As for the second joint, there is a slight reduction of 2.21%, resulting in a value of 0.310. Lastly, the third joint shows a small enhancement of 4.76%, going from 0.021 to 0.020. This study focused on developing a method for controlling the motion of the Robogymnast system by synchronising the stepper motors. To analyse the system's performance, a simulation was created using MATLAB/Simscape, examining various phases. The simulation and practical implementation of the controllers were carried out using MATLAB® and the STM32F microcontroller. In summary, this research contributes to the field of robotic systems by highlighting the significance of advanced control techniques and optimisation algorithms.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Engineering
Uncontrolled Keywords:	1). Robogymnast 2). Three link Robotic system 3). ACO Algorithm 4). Gravitational Search Algorithm 5). Acrobatic Robot 6). PID Control
Date of First Compliant Deposit:	22 November 2023
Last Modified:	22 Nov 2024 02:30
URI:	https://orca.cardiff.ac.uk/id/eprint/164126

Actions (repository staff only)

Edit Item