Following are the various modules of fuzzy controller:
- Rule base
- Data base
- Inference Engine
- It creates a normalized universe if discourse (domain) and performs a scale transformation i.e. if the domain is integer and the input value is float it converts it into integer. But if all the values belong to the same domain then there is no need of normalization.
- It converts the crisp normalized values to fuzzy values.
- Example: dirt-SD, MD, LD (Small Dirt, Medium Dirt, Large Dirt)
- A generalized way of representing a fuzzy set is using overlapped triangular membership function
3. Rule base:
- It decides all the rules that are necessary foor the calculation of the output it consist if all possible if..then rules. It combines various condition.
4. Data base:
- It provides and stroe the necessary information database is connected to all the modules thus, it can store input rules and output.
5. Inference engine:
- In inference engine all the rules are stored and based on the database and a particular input only a single rule is fired. Therefore it infers output based on rule base and database
- It converts the fuzzy values to a single point wise crisp value.
- Example: If the ouput is VS (very small) then it is defuzzified to crisp value i.e. 2 min [out of 0-60 min range]
- It maps the point wise crisp values to its physical domain. i.e. if the values was float in normalization module then it is converted to its required domain. But if normalization is not used then even denormalization is not required.