Next: About this document ...
- Hardware:
- A modified commercial robot (``rug warrior'',
diameter) was used which is driven by two DC motors. Average speed was
adjusted to
. Mechanical contact is detected by three
microswitches
in triangular configuration. Visual
information is aquired with two active infrared range finders
. The detection range was adjusted to
.
Interfacing between robot and computer is done tethered via a
standard parallel I/O card and an industry A/D-D/A card. The C++
program ran on a Linux system in the real-time mode.
- Sensor characteristics:
- Sensor signals are band-pass filtered
by the transfer function
. The transfer function in the LAPLACE domain is defined as
where
,
. The
output functions of the band-pass filters are denoted as
and
normalised to one. The band-pass characteristics of all collision
sensors
are identical and set to
and
. The signal
from each vision sensor is fed in parallel into a filter bank of ten
band-pass filters which frequencies are set to
;
is set to
,
throughout.
- Neuronal Circuitry:
- The robot has two neurons; one which
controls the speed
, the other the steering angle
. Normal
operation is straight forward motion (
). Both
neurons receive inputs from all sensors in a direct feed-forward
connectivity.
- Unconditioned retraction reaction:
- The unconditioned retraction
reaction uses only the collision sensor signals. These signals drive the
output neurons in such a way that an avoidance movement with a motion
vector pointing away from the site of stimulation is elicited.
- Learning:
- All band-pass filter outputs
from the
collision- and the vision sensors converge onto both neurons
where they are summed according to their synaptic weights
. Only the synaptic weights from the vision-sensors are
allowed to change starting from zero. The change of the weights is
achieved by the learning rule defined as:
where
is either
or
. The
constant
is set to
.
- Neuronal Output:
- The output of the neurons is defined as:
and
. The
asterisk denotes a convolution operation. The variables
and
represent the total sum of all learned contributions that
converge onto the
- and
-neuron respectively. The synaptic
weights in unconditioned reaction are kept constant at
and
.
Next: About this document ...
Bernd Porr
2001-07-16