下載/瀏覽Download
Download
Report
Transcript 下載/瀏覽Download
FORCE CONTROL STRATEGIES
WHILE DRIVING ELECTRIC
POWERED WHEELCHAIRS WITH
ISOMETRIC AND
MOVEMENT-SENSING JOYSTICKS
Outline
Introduction
Methods
Experimental Setup
Subjects
Data Collection
Statistical Analysis
Results
Conclusion
Introduction
Introduction
Approximately 220,000 Americans use
electric powered wheelchairs (EPWs) for
independent mobility.
An additional 125,000 Americans with
disabilities desire power mobility but cannot
functionally drive an EPW.
Introduction
Isometric joysticks (IJs)
rigid post : theoretically reduces the amount of
dexterity required for control
Movement sensing joysticks (MSJs)
Isometric joysticks
source : http://www.wheelchairnet.org/
Introduction
Compared Human Engineering Research
Laboratories (HERL) IJ with two control
functions to a conventional MSJ (PML
Flightlink, Ltd).
Subjects : experienced EPW users with upper
limb impairments from a variety of etiologies
Introduction
Sophisticated signal processing
statistically indistinguishable from the MSJ
basic control function
lower performance than using the MSJ
Regardless of the control software installed,
users tended to exert more force than
necessary on the IJ
Methods
Experimental Setup
Fitts’ Law
The model is based on time and distance
nine circular, black vinyl targets 155 cm in
diameter to the floor in a double semicircular
array
Experimental Setup
Distance of targets from the starting position
near targets : 305 cm
far targets : 538.5 cm
Mounted one of two test joysticks to a
Quickie P300 EPW (Sunrise Medical, 1994)
Experimental Setup
Force sensing algorithm (FSA)
program the IJ to operate as a simple IJ
Variable gain algorithm (VGA)
program the IJ to emulate an MSJ more closely
dead zone : low amplitude inputs ( 0.9 N)
occurring from unintentional/resting movements
produces no output
Experimental Setup
Subjects
Used an EPW with hand operated MSJ as the
primary means of mobility at least 20 h per
week
Ages
between 18 and 80 years
Tolerate testing for 2.5 h
Drive to each of nine targets three times with
each of three joystick
Subjects
Subject numbers in each diagnostic category :
4 (36.4%) with Cerebral Palsy (CP)
2 (18.2%) with traumatic brain injury (TBI)
2 (18.2%) with spinal cord injury (SCI)
1 (9.1%) with Muscular Dystrophy
1 (9.1%) with Spina Bifida (SB)
1 (9.1%) with Polio
Cerebral Palsy (CP) : 腦性麻痺
spinal cord injury (SCI)脊髓損傷
Spina Bifida (SB) :脊柱分裂
Traumatic brain injury(TBI) :創傷性腦損傷
Muscular Dystrophy :肌肉萎縮症
Polio :小兒麻痺疫
Data Collection
Accuracy
successful trial
halted the EPW within the target for at least 2 s
with a total trial time of no more than 40 s
Driving accuracy (DA)
the percentage of successful trials completed with
each joystick
Data Collection
Transducers
forward and reverse
“speed” axis
left and right
“direction” axis
Subjects
Data Collection
Average applied force
each trial by dividing the area under the force time
curve by the total trial time
Excess force
Each trial as the average of the differences
between the actual applied force
control efficiency (CE)
the Newton ‧ seconds (N‧s ) expended within the
operational range divided by the total N‧s of each
trial
Statistical Analysis
Alpha at values equal to 0.05 a priori
Used SPSS for univariate analyses and
nonparametric analyses
nonparametric statistics(Friedman test)
Average applied force and trial time were not
normally distributed
learning effect for average applied force
subsequent trials
nonparametric statistics(Mann–Whitney tests)
Determine if average applied force for each trial was
related to DA
Statistical Analysis
Spearman Rho correlations
evaluate the relationship between average applied
force and trial time for successfully acquired
targets
Results
Results
Average age
37.8 ± 10.9 years
Gender
male : 6 (54.6%)
female : 5 (45.5%)
Race
Caucasian : 6 (54.6%)
African-American : 4 (36.4%)
Asian-American : 1(9.1%).
Results
Results
No learning effect seen for average applied forc
any joystick
FSA p = 0.184, VGA p = 0.117, MSJ p = 0.804
Average applied force negatively correlated with
trial time for all three joysticks for successfully
acquired near and far targets
Average applied force was not related to DA for
any of the joysticks
FSA p = 0.306, VGA p = 0.126, MSJ p = 0.304
Results
Conclusion
Conclusion
The HERL IJ is a potential alternative control
interface for mobility for many individuals
with disabilities.
The IJ required more force to operate than
the MSJ, but subjects’ driving performance
was not affected.
Thank You for Your Attention!