Date of Award

August 2021

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Engineering

First Advisor

Anoop K Dhingra

Committee Members

WooJin Chang, Benjamin Church, Devendra Misra, Yongjin Sung

Keywords

Adaptive control, Computed torque control, Human lower extremity rehabilitation robot, LQR control, Rehabilitation robotics, Sliding mode control

Abstract

The World Health Organization reports that worldwide about 1 billion people have some form ofdisability. Of these, 110-190 million people have significant difficulties in functioning (mainly upper and lower extremity disability) independently. The major causes of human lower extremity disability include stroke, trauma, spinal cord injuries, and muscular dystrophy. Every 40 seconds, someone in the United States has a stroke. A statistic shows that approximately 65% of post-stroke patients suffer lower extremity impairment. Rehabilitation programs are the main method to promote functional recovery in disabled individuals. The conventional therapeutic approach requires a long commitment from a therapist or a clinician. Unfortunately, there is a persistent shortage of qualified physiotherapists/clinicians both in developing and developed countries. For therapists, the treatment process usually requires many hours with each patient, and the amount of such cases is constantly increasing. Therefore, an alternative to conventional treatment options is essential. Exoskeleton robot-assisted physical therapy has recently been studied extensively due to its proven effectiveness in providing different forms of physical therapy at any stage of physical recovery. To help disabled individuals cut down their disability period and regaining mobility quickly, we started development of a human lower extremity exoskeleton robot for physical therapy that can provide therapy during all stages of physical recovery. The development of an exoskeleton robot involves anatomic and anthropometric analysis of limbs, mechanical design, electrical and embedded control electronics system design, communication system design, and control algorithm design and realization. In this thesis, the human lower extremity anatomic and anthropometric analysis is presented from the perspective of a rehabilitation robot development. The kinematic and dynamic modeling of the robot are presented first. The Denavit-Hartenberg approach is used for kinematic modeling of the lower limbs. Both Newton-Eulers iterative approach and Lagrange energy methods are used for dynamic modeling. The LuGre friction model is used for simulating joint friction. A total of 7 linear and nonlinear control techniques (PID control, Computed torque control, Linear Quadratic Regulator, Model reference computed torque control, Adaptive control, Sliding mode control, and a Sliding mode control with a super twisting algorithm) are developed for the robot. 8 novel mechanisms are developed for providing all-natural movements for 7 active and 1 passive degree of freedom. The mechanical manipulator is designed and simulated in the CAD environment and fabricated using CAM processes. A real-time robot controller is developed to execute the control algorithms and interface the sensors and actuators. A graphical user interface was developed to interact with the robot and collect the data. Finally, recommendations for future research are presented.

Download

Share

COinS