Formation of a dynamic knowledge base of fuzzy inference systems for estimating changing in time objects
The article is a continuation of a series of author`s papers of dynamic fuzzy sets, which the main purpose is to process various information about the objects of complex systems represented by dynamic fuzzy descriptions. The factors that influence the change of fuzzy set in time and its membership f...
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | article |
| Lenguaje: | RU |
| Publicado: |
State University of Management
2019
|
| Materias: | |
| Acceso en línea: | https://doaj.org/article/ef703adc027d47358f19ee177461c6fa |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: | The article is a continuation of a series of author`s papers of dynamic fuzzy sets, which the main purpose is to process various information about the objects of complex systems represented by dynamic fuzzy descriptions. The factors that influence the change of fuzzy set in time and its membership function, for example, the change of the range of included values of variables in the fuzzy set leads to set for the “old” value of the variable a “new” value of the membership function; transformation of the type of membership function are allocated. The decision-making algorithm based on the fuzzy inference method for estimating objects changing in time is considered. The concept of dynamic membership, dynamic rules and dynamic rule base are determined. We propose to add the process of fuzzy inference with the stage of forming a dynamic rule base, which is able to be updated when the behavior of the dynamic object under study changes, meanwhile, to add and remove rules as a whole, and to change any part of the rule is allowed. The conditions, on the basis of which, occurs a change in fuzzy rules are determined. The process of formation of a linguistic variable in the construction of interval estimates of dynamic objects, which consists of six stages and includes the representation and ordering of known expert control points; search for missing values by interpolation and approximation; determining the number of terms of the linguistic variable and their boundaries and the formation of dynamic term-set of values of indicators is considered. In the same way as the dynamic membership function constitutes a fuzzy surface in “the degree of membership-parameter-time space”, a linguistic variable composed of dynamic terms, represents a graphically broken surface with n - vertices according to the number of terms. |
|---|