Page 149 - Robots Androids and Animatrons : 12 Incredible Projects You Can Build

Fuzzy logic
In 1965, Lotfi Zadah, a Professor at the University of California at
Berkeley, first published a paper on fuzzy logic. Since its inception,
fuzzy logic has been both hyped and criticized.
In essence, fuzzy logic attempts to mimic in computers the way
people apply logic in grouping and feature determination. A few
examples should clear this “fuzzy” definition. For instance, how is a
warm, sunny day determined not to be warm but to be hot instead,
and by whom? The threshold of when someone considers a warm
day hot depends on a person’s personal heat threshold and the
influence of his or her environment (see Fig. 6.27).
There is no universal thermometer that states at 81.9 degrees
Fahrenheit (°F) it is warm and at 82°F it is hot. Extending this
example further, a group of people living in Alaska has a different
set of temperature values for hot days when compared to a group
of people living in New York, and both these values will be different
from that of a group of people living in Florida. And let’s not forget
seasonal variations. A hot day is a different temperature in winter
than summer. So what this boils down to is that classifications (for
example, “What is a hot day?”) may be a range of temperatures
128 determined by the opinions of a group of people. Further classifi-
cations can be differentiated by different groups of people.
Any particular temperature will find membership in the group where
it fits into the range of values. Sometimes a temperature will fit into
two overlapping groups. True membership will then be determined
by how a particular temperature varies from the median values.
The idea of group and range classifications can be applied to many
other things, like navigation, speed, and height. Let’s use height for

6.27 Grading temperature from warm to hot, gradually or by step
Team LRN
Chapter six