Control is a mechanism used to guide or regulate the operation of a system.
Feedback Control is a mechanism for guiding or regulating a system by returning a fraction of the output to the input.
The signal returned to the input from the output in feedback control can be called the Error Signal.
In classical operation of process control one has to solve the non-linear equation u for the output, and also must have found the constants involved in the equation u. The function may be unknown or difficult to obtain.
In a fuzzy controller, the non-linear equation for the process is represented by a fuzzy mapping
A Fuzzy Controller is a device that models a vaguely described process.
There are two types of fuzzy controller: Mamdani Controllers and Takagi-Sugeno-Kang Controllers.
Feedback mode is where the fuzzy controller is used for controlling.
Feedforward mode is where the fuzzy controller is used for predicting.
All inputs to, and outputs from, a fuzzy controller are in the form of linguistic variables.
Regulation is the process of keeping the output close to a set reference point.
Coarse is a synonym of rough.
A clever way of representing a rule base where the linguistic variables have large term sets is to put them in a table. Consider SPEED x SIZE -> THREAT:
| SLOW | AVERAGE | FAST | |
|---|---|---|---|
| SMALL | FRIENDLY | FRIENDLY | ALRIGHT |
| MEDIUM | FRIENDLY | ALRIGHT | ALRIGHT |
| LARGE | ALRIGHT | DANGEROUS | DANGEROUS |
As the term sets get bigger, it makes sense to put them in a table like above with single letters representing the words and some of the entries blank.