Optimal design and positioning control performance of a 2-DOF robotic finger

Mohamad Yuden, Mohamad Adzeem (2018) Optimal design and positioning control performance of a 2-DOF robotic finger. Masters thesis, Universiti Teknikal Malaysia Melaka.

	Text (24 Pages) Optimal Design And Positioning Control Performance Of A 2-DOF Robotic Finger - Mohamad Adzeem Mohamad Yuden - 24 Pages.pdf - Submitted Version Download (360kB)
	Text (Full Text) Optimal design and positioning control performance of a 2-DOF robotic finger.pdf - Submitted Version Restricted to Registered users only Download (10MB)

Abstract

This research focuses on the positioning control performances of two degrees of freedom (2-DOF) robotic finger mechanism in achieving precision motion control. The research outcomes were expected to contribute in wider robotic hands for precision applications and suggest that the advantages of 2-DOF robotic finger can be carried over to precision and dexterous tasks. The research investigates the design of a 2-DOF robotic finger mechanism and its control strategies for achieving high precision grasping as initial research towards developing a multi-fingered robotic hand system. Behaviors such as instability, large steady-state error and poor transient performance often occurred in the robotic hand mechanism. Therefore, the goal of this research is to design a 2-DOF robotic finger mechanism and compare the performances of several controllers for positioning motion control and evaluate the effectiveness of controllers by Point-to-Point (PTP) control and tracking control. In this research, the proposed controllers will depend on the angular position control of each motor joints, i.e. the position control of the 2-DOF robotic finger mechanism. In order to achieve the research objectives, the research was implemented in three (3) main phases. Phase 1 involve the optimization of the robotic hand design using Finite Element Analysis (FEA) using Solidworks software and its experimental setup. In Phase 2, the robotic finger mechanism mathematical modeling and system identification methods were discussed and compared. In phase 3, two categories of control system experiments were carried out which are the open-loop control system and the closed-loop control system. For open-loop system, the evaluation was done using the step input signal. Meanwhile, the closed-loop system was carried out for uncompensated and compensated system using several reference angles. Three different control startegies namely (i) Proportional Integral Derivative (PID) controller (ii) Fuzzy Logic controller (FLC) and (iii) Linear Quadratic Regulator (LQR) controller were chosen to be compared via simulation and experimental works. The controller results were validated by Point-to-Point (PTP) control and tracking control with frequency from 0.1 Hz to 0.5 Hz at different reference amplitudes. From the analyze results, it is proven that the Fuzzy controller gave the best performances and has higher level of adaptability for the PTP control with improvements in both response time by 97.9 % (0.075 s) and steady-state error by 99.5 % (0.01 °) in compared to the uncompensated system. Meanwhile, it was concluded that LQR controller exhibits the best tracking control performances. The LQR controller had demonstrated an improvement in steady-state error by 98.5 % (0.11 °) over the uncompensated system in a series of experimental tracking tests. It was also concluded that the 2-DOF robotic finger mechanism had also succeeded the grasping tasks with the specific reference trajectory using the Fuzzy controller.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Robots Control systems, Digital control systems, Robotics, Robot hands, A 2-DOF Robotic Finger
Subjects:	T Technology > T Technology (General) T Technology > TJ Mechanical engineering and machinery
Divisions:	Library > Tesis > FKE
Depositing User:	Nor Aini Md. Jali
Date Deposited:	28 Aug 2019 08:50
Last Modified:	13 May 2022 11:50
URI:	http://eprints.utem.edu.my/id/eprint/23486
Statistic Details:	View Download Statistic

Actions (login required)

View Item