Graduation Year

2009

Document Type

Thesis

Degree

M.S.B.E.

Degree Granting Department

Biomedical Engineering

Major Professor

Rajiv Dubey, Ph.D.

Committee Member

Redwan Alqasemi, Ph.D.

Committee Member

Kyle Reed, Ph.D.

Keywords

rehabilitation robotics, dual-trajectory, mobile robot, manipulator, redundancy, adl

Abstract

The main objective of my research is to improve the control structure for the new Wheelchair Mounted Robotic Arm (WMRA) to include new algorithms for optimized task execution; that is, making the WMRA a modular task oriented mobile manipulator. The main criterion to be optimized is the fashion in which the wheelchair approaches a final target as well as the starting and final orientation of the wheelchair. This is a novel approach in non-holonomic wheeled manipulators that will help in autonomously executing complex activities of daily living (ADL) tasks.

The WMRA is a 9 degree of freedom system, which provides 3 degrees of kinematic redundancy. A single control structure is used to control the WMRA system, which gives much more flexibility to the system. The combination of mobility and manipulation expands the workspace that a mobile base attains to a manipulator. This approach opens a broad field of applications: from maintenance and storage to rehabilitation robotics. This structure is based on optimization algorithms that can resolve redundancy based on several subtasks: maximizing the manipulability measure, minimizing the joint velocities (hence minimizing the energy), and avoiding joint limits. This work utilizes redundancy to control 2 separate trajectories, a primary trajectory for the end-effector and an optimized secondary trajectory for the wheelchair. Even though this work presents results and implementation in the WMRA system, this approach offers expandability to many wheeled base mobile manipulators in different types of applications.

The WMRA usage was simulated in a virtual environment, by developing a test setting for sensors and task performance. The different trajectories and tasks can be shown in a virtual world created not only for illustration purposes, but to provide training to the users once the system is ready for use.

Share

COinS