Information Technology Reference
In-Depth Information
0.5
50
motor1
motor2
motor3
motor4
inflammation
0.45
45
0.4
40
0.35
35
0.3
30
0.25
25
0.2
20
0.15
15
0.1
10
0.05
5
0
0
0
500
1000
1500
2000
2500
3000
time
Fig. 6.
Effect of the current change on the inflammation level
Figure 3 shows the SOM as generated using the state input feature vectors
taken from the above experiments. A clear cluster, representing the homeostatic
state, can be seen in the upper left region of the table. This cluster comprises
of states which contain value 0 for all TLR responses. This value represents no
triggering activity of the TLRs. By contrast, the dark region in the lower right
quadrant represents triggering of TLRs both to switch on a fan and to switch two
motors off to prevent damage. The region in the centre of the bottom row rep-
resents triggering of a single fan, and is bordered by regions to the left and right
which represent switching on the fan in the other motor compartment (left) and
switching off a motor in the overheating compartment (right). These adjacent
regions can be used to highlight what might happen if inflammation caused by
the single fan in the first motor compartment persists and is required to spread
through the SOM. Activation of the adjacent regions mentioned will trigger
preventative high-level actions appropriate for these closely related states. For
example reducing current in the affected motors is likely to be one of the actions
taken in order to pre-empt the triggering of the TLRs in the other components.
6Con lu on
A scheme for incorporating low-level damage prevention and maintenance activities
into a coherent biologically inspired control paradigm has been proposed, based on
an innate immune system. Three important aspects of the innate immune
