Date of Publication

3-2011

Document Type

Master's Thesis

Degree Name

Master of Science in Manufacturing Engineering

Subject Categories

Manufacturing | Robotics

College

Gokongwei College of Engineering

Department/Unit

Manufacturing Engineering and Management

Thesis Adviser

Marlon Luis M. Musngi

Defense Panel Chair

Homer S. Co

Defense Panel Member

Edwin J. Calilung
Arthur Pius P. Santiago

Abstract/Summary

Model-based control is now a significant technology for the control of robots. Models and control schemes are continuously refined to meet the requirements of higher performance and lower cost. The control strategies used in most robots involve position coordination in the Cartesian space through the inverse kinematics method. Inverse kinematics comprises the computations to determine the joint angles needed to achieve the position and orientation for the robot end-effector. The inverse kinematics problem is usually complex for robotic manipulators. There are three traditional methods used for solving inverse kinematics problems: geometric, algebraic and iterative. Computing for the inverse kinematics solution using these traditional methods is a time-consuming study, especially when the joint structure of the manipulator is more complex. The computation of inverse kinematics using artificial neural networks is particularly useful where less computation time is needed, such as in real time adaptive robot control. Traditional methods will become prohibitive due to the high complexity of the mathematical structure of the formulation, wherein robots have to work in the real world that cannot be modeled concisely using mathematical expressions. A neural network-based inverse kinematics solution methods yield multiple and precise solutions with an acceptable error and it is suitable for real-time adaptive control of robotic manipulators. This research focuses on the design and development a control system for a 5 DOF revolute robot arm, using the Elman neural network based inverse kinematics solution approach. From the recurrent networks family, Elman Network is selected because of its feedback loops that have a weighty impact on the learning capability and performance of the network. This research integrates the mechanical and electronic systems design, with the model-based control algorithm, to establish an integral robot control system.

Abstract Format

html

Language

English

Format

Electronic

Accession Number

CDTG004972

Shelf Location

Archives, The Learning Commons, 12F, Henry Sy Sr. Hall

Physical Description

1 computer optical disc, 4 3/4 in.

Keywords

Robots—Kinematics; Robots—Control systems

Upload Full Text

wf_yes

Embargo Period

5-5-2022

Download

Share

COinS