
DESIGNING INTELLIGENT SYSTEMS MANAGEMENT TRANSPORT ENTERPRISES ENTROPY APPROACH
Author(s) -
Денис Юрійович Зубенко,
Alexsandr Kuznetzov
Publication year - 2016
Publication title -
eureka, physics and engineering./eureka, physics and engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.303
H-Index - 5
eISSN - 2461-4262
pISSN - 2461-4254
DOI - 10.21303/2461-4262.2016.00025
Subject(s) - complexity management , computer science , computational complexity theory , entropy (arrow of time) , complex system , theoretical computer science , artificial intelligence , algorithm , physics , quantum mechanics , marketing , business
Major construction this article discusses the methodological foundations of intelligent control systems for complex technical objects in the class of hierarchical control systems based on entropy approach. The notion of design space to MIS (management systems dynamically-changing facilities), its topology and metric. The generalized criterion of the complexity of the design decisions. The approaches to the definition of entropy at the executive level, the level of coordination and planning level. The problem of the comparative assessment of the complexity of the control algorithms synthesized based on neural network technology and fuzzy logic. As part of the proposed concept design ISU to the methodological principles of design to the ISU. On the basis of estimates of the complexity of design decisions on the levels of planning, coordination and executive level. This allows a common position to ensure the required quality of design decisions by limiting the complexity of their implementation. Too complex sentence. Split into simpler evaluation of the complexity of the proposed management and algorithmic complexity control algorithms based on entropy approach, which allows on the basis of scalar criterion to evaluate the complexity of integrated design solutions. A method for estimating the complexity of interpolation algorithms based on soft computing technology that allows for a reasonable complexity of algorithms for constructing nonlinear dynamic models, nonlinear regulators.