Democritus University of Thrace (DUTH)
Ανάλυση, σχεδίαση και εκπαίδευση ασαφών γνωστικών δικτύων με εφαρμογές στην προσέγγιση και στον έλεγχο συστημάτων
Abstract
dc:descriptionFuzzy cognitive maps (FCM) are conclusion sets that use cyclically-directed graphs that represent the causal relationships between concepts. They use a symbolic representation to describe and model a physical system. A fuzzy cognitive map consists of nodes, each representing a feature of the system, while interactions between nodes represent the dynamics of the system. An FCM incorporates accumulated experience and knowledge about the system's operation as a result of its method of construction, ie by using experts whoknow the system's operation and behavior.In this thesis, an attempt is made to expand the FCM to support the close interaction with the physical system they describe, and thus become suitable for control and decision-making applications. The proposed expansion differs from the traditional FCM and is now called the Fuzzy Cognitive Network (FCN).We study the existence of their stable equilibrium points using a properly defined contrac- tion mapping theorem. It turns out that when the weights of the interfaces meet certain conditions that are related to the FCN size, the FCN node values will converge to a unique solution regardless of their original states, otherwise if the weights do not meet the specified conditions The existence or uniqueness of the balance can not be guaranteed. As a consequence of these results, the rules of the form "If weigh then equilibrium point" are more appropriate to represent the behavior of a FCN.
Degree
thesis:*- Grantor dc:publisher
- Democritus University of Thrace (DUTH)
- Year dc:date
- 2017
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kottas, Theodoros
- Κώττας, Θεόδωρος
Subjects
dc:subject × 6Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/41264