Metodologia de controle preditivo baseado em modelo Fuzzy evolutivo

• Ano de defesa: 2018
• Autor(a) principal: AZEVEDO JÚNIOR, Arnaldo Pinheiro de
• Orientador(a): SERRA, Ginalber Luiz de Oliveira
• Banca de defesa: SERRA, Ginalber Luiz de Oliveira , SOUZA, Francisco das Chagas de , BARRETO, Gilmar , ROCHA FILHO, Orlando Donato
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Maranhão
• Programa de Pós-Graduação: PROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA DE ELETRICIDADE/CCET
• Departamento: DEPARTAMENTO DE ENGENHARIA DA ELETRICIDADE/CCET
• País: Brasil
• Palavras-chave em Português: Controle preditivo; Modelagem fuzzy evolutiva; Pratical nonlinear model predictive control
• Palavras-chave em Inglês: Predictive control; Evolving fuzzy modeling; Pratical nonlinear model predictive control
• Área do conhecimento CNPq: Eletrônica Industrial, Sistemas e Controles Eletrônicos
• Link de acesso: https://tedebc.ufma.br/jspui/handle/tede/2418

The objective of this work is to propose a methodology based on the combination of predictive control and evolving fuzzy modeling. Predictive control is an advanced industrial technique, capable of calculating the control signal applied to the process from a prediction of its future behavior. Evolving fuzzy modeling is a model identification technique, capable of acquisition of Knowledge of the process in the form of IF-THEN fuzzy rules, as well as evolving its structure and updating its parameters. This work proposes a predictive control methodology based on an evolving fuzzy model capable of controlling multivariable processes with nonlinear dynamics. The predictive control technique used is the Practical Nonlinear Model Predictive Control, which calculates the control signal from an approximation of the non-linear prediction model of the process to be controlled. The prediction model used is obtained from an evolving version of the Gustafson-Kessel fuzzy clustering technique and the least squares recursive algorithm. The proposed controller is able to improve its tracking capabilitie of a reference trajectory, because, it evolves the structure of the non-linear prediction model from the extraction of dynamic knowledge of the inputs and outputs of the process to be controlled. In order to evaluate the proposed methodology, it was applied to the control of three non-linear benchmarking processes known in the literature.