Computer World. Матросова Т.А. - 74 стр.

UptoLike

ВУЗ: 

УлГТУ | Ульяновск

Составители: 

Рубрика: 

73
Fuzzy Perception
A fuzzy perception is an assessment of physical condition that is not measured
with precision, but is assigned an intuitive value. In fact, the fuzzy logic people assert
everything in the universe is a little fuzzy, no matter how good your measuring
equipment is. It will be seen below that fuzzy perceptions can serve as a basis for
processing and analysis in a fuzzy logic control system.
Measured, non-fuzzy data is the primary input for the fuzzy logic method.
Examples: temperature measured by a temperature transducer, motor speed,
economic data, financial markets data, etc. It would not be usual in an electro-
mechanical control system or a financial or economic analysis system, but humans
with their fuzzy perceptions could also provide input.
In the fuzzy logic literature, you will see the term «fuzzy set.» A fuzzy set is a
group of anything that cannot be precisely defined. Consider the fuzzy set of «old
houses.» How old is an old house? Where is the dividing line between new houses
and old houses? Is a fifteen year old house an old house? How about 40 years? What
about 39.9 years? The assessment is in the eyes of the beholder.
Other examples of fuzzy sets are: tall women, short men, warm days, high
pressure gas, small crowd, medium viscosity, hot shower water, etc.
When humans are the basis for an analysis, we must have a way to assign some
rational value to intuitive assessments of individual elements of a fuzzy set. We must
translate from human fuzziness to numbers that can be used by a computer. We do
this by assigning assessment of conditions a value from zero to 1.0. For «how hot the
room is» the human might rate it at .2 if the temperature were below freezing, and the
human might rate the room at 9, or even 10, if it is a hot day in summer with the air
conditioner off.
You can see these perceptions are fuzzy, just intuitive assessments, not precisely
measured facts.
By making fuzzy evaluations, with zero at the bottom of the scale and 1.0 at the
top, we have a basis for analysis rules for the fuzzy logic method, and we can
accomplish our analys is or control project. The results seem to turn out well for
complex systems or systems where human experience is the only base from which to
proceed, certainly better than doing nothing at all, which is where we would be if
unwilling to proceed with fuzzy rules.
TEXT 2
DESI GN OF A BITMAPPED MULTILI NGUAL WORKSTATION
Providing computer support for English text is simpler in several important
respects than supporting other natural languages. Most English words require no
diacritical marks. Internal storage of ASCII codes allows implicit collation of terms.
Display of English text requires only upper- and lowercase letters to generate output
comparable to printing generated by other means.
By contrast, many non-English languages use characters that increase the
complexity of computerization. Langauges that use diacritical marks for certain

Страницы