| MATLAB Function Reference | Search  Help Desk |
| taksug | Methods |
class for implementing the taksug fuzzy functions
Syntax
m=taksug(varargin)
m=taksug(taksug_obj)
m=taksug(structure)
m=taksug(name, n_inputs, n_outputs)
Description
m=taksug(varargin)
returns an empty taksug object.
m=taksug(taksug_obj)
returns a copy of the taksug_obj object.
m=taksug(structure)
initializes a new taksug object
with the appropriate fields set according to the values
defined in the fields of structure.
m=taksug(name, n_inputs, n_outputs)
initializes a new taksug
object called name, with number of inputs n_inputs and
number of outputs n_outputs.
Remarks
The taksug class is a child of the mapping class and allows the definition of a static system using a fuzzy rule based Takagi-Sugeno (TS) representation. The class inherits all the fields of the mapping class and defined a series of new attibutes required for implementing the TS representation of a model. At the top level the following fields are defined:n_rules defines the number of fuzzy rules of the model
model_code defines the structure of the fuzzy rules
centers position of the center of each rule
ivariances inverse of the projected covariance matrix
of the cluster associated to each rule. In other words,
it is the quadratic matrix defining the local metric (or
the shape) of the rule
linears hyperplane associated to each rule (consequence of the rule)
m index of the fuzzyness of the model
mfs description of the fuzzy sets while they are defined along each dimension.
rls description of the rules base.