Arobust control design for a robot manipulator

Dissertant

Ibrahim, Muhannad Zuhayr

University

University of Technology

Faculty

-

Department

Department of Control and Systems Engineering

University Country

Iraq

Degree

Master

Degree Date

2006

English Abstract

Early robot manipulator operates with simple controllers and does not take the advantages of computer in moving the manipulator to track a desired trajectory with a small error in tracking.

On the other hand, modern controller has been developed during the last two decades.

The classification of these controllers is between those based on soft computing techniques or those based on optimal robust theory.

In this thesis the optima! robust control theory is adopted for controlling the motion of n-joint link manipulator in the environments of parameters uncertainty and input stochastic disturbance existence.

A dynamic model of SCARA robot manipulator with different level of uncertainty in load parameters is considered as a first case study robust design approach.

A robust controller based on LQR theory is designed to ensure good tracking and fast settling time.

Such design shows many-drawbacks.

especially in selecting the uncertainty parameters.

A second robust controller based on minim ax approach is designed.

The idea is to minimize the optimal control input (torque) and maximize the attenuation of the coupled uncertainty.

A necessary and sufficient condition is obtained for a robust optimal control design for a general n-joints link robot manipulator.

Simulation of 2 and 3 joints are performed to test two main features.

They track different desired trajectories as well as standing against parameter changes or / and stochastically input disturbances.

To assure the validity of minim ax approach irrespective of the number of joints, a three-joint link manipulator under the same conditions and for different tracking trajectory requirements is simulated.

The results show the applicability of the minim ax optimal solution and the performance is well accepted without enlarging the required torque.

Main Subjects

Mechanical Engineering

Topics

• Computing
• Control systems
• Robots

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-305600