Built-in Fuzzy Sets
|
Home
NEW:
US Markets Daily |
Built-in Fuzzy SetsFuzzy sets are one of the 3 components of our FL controllers, along with the Variable Set and the Rule Set. The controller being constructed with a given shape (3x3 or 5x5), fuzzy sets will obviously have to conform the type of FL controller being built. Again, in order to keep sFLC3 simple to use, a number of predefined fuzzy sets have been built in and can therefore be called and modified with just one or 2 API calls (see sample code for more details). Lastly, the semantics behind variables has not been ignored and has also been kept simple by attaching a given set of "labels" to each of our fuzzy states. Labels can be changed like anything else. const string strStates5[3][5] ={ "Very Low", "Low", "Med", "High", "Very High", "Very Neg.", "Neg.", "Null", "Pos.", "Very Pos.", "Very Small", "Small", "Avg.", "Big", "Very Big" }; const string strStates3[3][3] ={ "Neg.", "Null", "Pos.", "Small", "Avg.", "Big" };
"3x3 Shape"The 3x3 Shape means that the 2 variables will both use a 3-state fuzzy set
(such as "Low", "Medium", and "High"). The sFLC3 API includes 9 predefined
fuzzy set models to facilitate the building of fuzzy logic controllers: To better comprehend what is behind those names, a spreadsheet here describes and displays the different shapes: FS3.XLS or FS3.MHT Nota Bene 1: All fuzzy sets are normalised to the [-1:+1] range, meaning -1 is the 'absolute' definition of "Very Low" or "Very Small", and +1 is obviously the opposite "Very High" or "Very Big". Secondly, a variable associated to a fuzzy set will be given a 'degree of membership' for each Fuzzy State in the associated Fuzzy Set. That degree of memebership is measured from 0 to 1 or 0% (not attributable to that state) to 100% (fully attributable to that state). Nota Bene 2: Users are limited to predefined Fuzzy Set Models and can also create their own custom fuzzy sets.
"5x5 Shape"Again, this means that both variables will use a 5-state fuzzy set. The sFLC3 API here includes 5 predefined fuzzy sets: "Regular","Reg-Wide", "Reg-Narrow", "Avg", "Wide" And here again, we have a spreadsheet available in 2 different formats to describe the shape and behaviour of those fuzzy sets: FS5.XLS and FS5.MHT
|
Page last modified:
May 08, 2008 |