0
Research Papers

Control of a Passive Mobility Assist Robot

[+] Author and Article Information
Ji-Chul Ryu

Mechanical Systems Laboratory, Department of Mechanical Engineering, University of Delaware, Newark, DE 19716jcryu@udel.edu

Kaustubh Pathak

Mechanical Systems Laboratory, Department of Mechanical Engineering, University of Delaware, Newark, DE 19716

Sunil K. Agrawal1

Mechanical Systems Laboratory, Department of Mechanical Engineering, University of Delaware, Newark, DE 19716agrawal@udel.edu

1

Corresponding author.

J. Med. Devices 2(1), 011002 (Mar 07, 2008) (7 pages) doi:10.1115/1.2889056 History: Received September 17, 2007; Revised February 05, 2008; Published March 07, 2008

In this paper, a control methodology for a mobility assist robot is presented. There are various types of robots that can help persons with disabilities. Among these, mobile robots can help to guide a subject from one place to the other. Broadly, the mobile guidance robots can be classified into active and passive types. From a user’s safety point of view, passive mobility assist robots are more desirable than the active robots. In this paper, a two-wheeled differentially driven mobile robot with a castor wheel is considered the assistive robot. The robot is made to have passive mobility characteristics by a specific choice of control law, which creates damperlike resistive forces on the wheels. The paper describes the dynamic model, the suggested control laws to achieve the passive behavior, and proof of concept experiments on a mobile robot at the University of Delaware. From a starting position, the assistive device guides the user to the goal in two phases. In the first phase, the user is guided to reach a goal position while pushing the robot through a handle attached to it. At the end of this first phase, the robot may not have the desired orientation. In the second phase, it is assumed that the user does not apply any further pushing force while the robot corrects the heading angle. A control algorithm is suggested for each phase. In the second phase, the desired heading angle is achieved at the cost of deviations from the final position. This excursion from the goal position is minimized by the controller. This control scheme is first verified in computer simulation. Then, it is implemented on a laboratory system that simulates a person pushing the robot, and the experimental results are presented.

FIGURES IN THIS ARTICLE
<>
Copyright © 2008 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Figure 4

The robot’s positions, xc and yc, plotted individually. The robot is guiding a user along the S-shaped curve shown in Fig. 3. Phase II begins at the time t=129.7s. In Phase II, xc and yc go through an excursion from the goal.

Grahic Jump Location
Figure 5

The desired and actual heading angle plots. Discontinuities in the desired heading angle indicate that the robot reaches waypoints and goes toward the next waypoint. At t=129.7s, Phase II starts.

Grahic Jump Location
Figure 6

Kr and Kl determined by the control algorithms in Phases I and II. In Phase I, until t=129.7s, the values of Kr and Kl show symmetry and are positive at all times as expected.

Grahic Jump Location
Figure 7

Experimental snapshots show the robot’s position and orientation on the way to the goal position from the start. Three waypoints were chosen so that the robot could avoid the obstacles in the environment.

Grahic Jump Location
Figure 8

The experimental result in which the robot passes through three waypoints between the start and the goal with the final heading orientation of 83.5deg.

Grahic Jump Location
Figure 1

A differentially driven mobile robot in Cartesian space described by three coordinates (xc,yc,θ)

Grahic Jump Location
Figure 2

Robot being steered, guided by the waypoints. At the position shown, the robot should ideally track the planned orientation to the next waypoint θp. However, if ∣θ−θp∣ is too big to satisfy Eq. 17, an intermediate orientation θd is chosen, which is close to θp while satisfying the inequality in Eq. 17.

Grahic Jump Location
Figure 3

Simulation result in which the robot executes an S-shaped curve with 23 waypoints between the start position and the goal with the desired heading direction of 180deg at the goal

Grahic Jump Location
Figure 9

The robot’s positions, xc and yc. Phase I is until t=18.2s, then xc and yc go through an excursion from the goal.

Grahic Jump Location
Figure 10

The desired and actual heading angles. Discontinuities in the desired heading angle indicate that the robot reaches a waypoint and goes toward next waypoint or the goal. When the robot is not fully oriented toward the next waypoint or the goal around it, the desired heading angle changes rapidly, which results in increase in error as shown in this graph. However, we can avoid this effect by considering that the robot arrives at each waypoint when the robot reaches within a circle of radius ϵ. At t=18.2s, Phase II occurs.

Grahic Jump Location
Figure 11

Kr and Kl, which are determined by the control algorithms in Phases I and II. In Phase I, until t=18.2s, the values of Kr and Kl show symmetry and are non-negatives at all times as expected.

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In