Rockwell-automation RSLogix 5000 Fuzzy Designer User Manual download pdf (Page 16)

Publication LOGIX-UM004A-EN-P - March 2007

16 Get Started with FuzzyDesigner

First, the numerical values of Temperature and Humidity get their

meaning. In our case, the current setting of the Temperature is such

that it is both 85% Medium and 40% High. Humidity is both 80% High

and 50% Very High. The first rule is thus 80% true for the current

inputs while the second rule is 40% true when using minimum for

the and operation. The first rule states that, if 100% satisfied, the

compressor should run at Medium speed. Currently, the first rule is

only 80% fulfilled, so one method of how to consider that the rule is

only 80% fulfilled is to truncate the Medium fuzzy set for the output at

the level 0.8.

A similar situation happens with the second rule where High

compressor speed is only 40% fulfilled. As both rules are used at the

same time, their conclusions must be combined to get a fuzzy value

for the output, which is compressor speed. The partially-fulfilled

Medium and High fuzzy sets are unified, and a single fuzzy value is

assigned to Compressor Speed. As conventional control systems

cannot deal with fuzzy values, the fuzzy instruction includes

conversion from a fuzzy to a crisp value. For this case, the center of

gravity for the green area is computed and used to represent the

original fuzzy value.

To summarize, the designer has to:

• define input and output variables.

• cover the interval of the respective variable by fuzzy sets (that is,

membership functions).

• write if-then rules using labels of the fuzzy sets defined

previously.

Potential Use of Fuzzy Logic

FuzzyDesigner enables you to enhance the functionality of existing or

new control and decision making systems in various branches of

industry.

The fuzzy system designed and generated by FuzzyDesigner can be

used in control systems, for example, as a direct nonlinear fuzzy-rule

based controller, PID-feedback control system supervisor, or a process

model in a Model Predictive Control scheme. Input and output filters

are used for signal preprocessing such as filtering, deriving trends, and

many other functions that might add dynamics to the static I/O map

generated from fuzzy rules. Input filters can also be designed in

FuzzyDesigner. Output filtering is an option and contains, for

instance, a discrete integrator fed by the output of the Fuzzy Add-On

Instruction.

1 2 ... 11 12 13 14 15 16 17 18 19 20 21 ... 155 156

No comments

Rockwell-automation RSLogix 5000 Fuzzy Designer User Manual Page 16