Information Technology Reference
In-Depth Information
concerning the adaptation to the desires of the inhabitants, and it processes the co-
stimulative signals that control the ventilation. The network is shown in figure 6.
In principle, the network consists of a number of functions that are closely
connected with each other such that there is a strong interdependence between them.
These functions control the buffer for the antigens and for the results, they compute
the affinities between antigens and antibodies and among antibodies, they determine
the mutation of antibodies and the production of new ones, they adapt the temperature
parameter according to the desires of the users, they eliminate useless elements, they
determine the key elements by a number of interrelated sub-functions, and finally
select the best suited antibody.
4 The Realization of the AIS
A prerequisite for the AIS is the encoding of the attribute values. Some of them are
real values, others are symbolic. For a uniform representation we decided to encode
them as binary strings. In order to do this the real type parameters are divided into a
finite number of intervals and their values are replaced with these intervals so that we
end up with only a finite domain for each parameter. The implementation has been
done in Java. For each main function and main component a class together with a
number of subclasses is defined. Each class has a function for the input of values and
an output function. The encoding is shown by the example of the codes for the rooms
and some important classes of the implementation are described in more detail in the
following subsections.
4.1 The Code of the Rooms as an Example for the Encoding
Each room in the house is encoded by a binary string according to the following
criteria:
•
Duration of use,
•
frequency of use,
•
preferred time of use,
•
preferred temperature.
For instance for the bathroom and the living room we fixed values according to
table 1:
Table 1.
Parameter values for the encoding of rooms
Duration of
use
Frequency of
use
Preferred time of
use
Preferred
Temperature
Bathroom
low
often
morning/evening
22 - 24 °C
Living Room
high
only once
evening
20 - 22 °C
All rooms are encoded according to this scheme. The values of the criteria
mentioned above are encoded in such a way that the codes of similar values are close
to each other. Table 2 gives an example for the frequency of use. This encoding has
