Algoritmo diferencial evolutivo aplicado à identificação de sistemas dinâmicos não-lineares

The main objective of this work is to contribute in the structure’s selection stage in system identification process using artificial intelligence techniques, especially the Differential Evolutionary (DE) algorithm. To reach this goal a new NARMAX representation is proposed where exponents with real...

Descripción completa

Detalles Bibliográficos
Autor: Morais, Josué Silva de
Tipo de recurso: tesis doctoral
Estado:Versión publicada
Fecha de publicación:2013
País:Brasil
Institución:Universidade Federal de Uberlândia (UFU)
Repositorio:Repositório Institucional da UFU
Idioma:portugués
OAI Identifier:oai:repositorio.ufu.br:123456789/18418
Acceso en línea:https://repositorio.ufu.br/handle/123456789/18418
Access Level:acceso abierto
Palabra clave:Engenharia elétrica
Sistemas não lineares
Identificação de sistemas
Métodos de simulação
Diferencial evolutivo
NARMAX expoente reais
Differential evolution
Identification systems
NARMAX real exponent
CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
Descripción
Sumario:The main objective of this work is to contribute in the structure’s selection stage in system identification process using artificial intelligence techniques, especially the Differential Evolutionary (DE) algorithm. To reach this goal a new NARMAX representation is proposed where exponents with real values are used. The evaluation of the proposed method was performed through an application developed specifically for this work with a QT package using C + + and parallel programming techniques. Different kinds of artificial systems were tested to evaluate the proposed methodology. Finally the methodology is applied to a dc-dc converter Boost -Buck and a biochemical polymer reactor and the results are presented.