| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 7.29 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Neste documento descreve-se o projecto desenvolvido na unidade curricular de Tese e Dissertação durante o 2º ano do Mestrado de Engenharia Electrotécnica e de Computadores no ramo de Automação e Sistemas, no Departamento de Engenharia Electrotécnica (DEE) do Instituto Superior de Engenharia do Porto (ISEP).
Com a revolução dos sistemas automóveis, toda a gestão do funcionamento do motor passou a estar a cargo de um sistema electrónico, originando a gestão electrónica do motor. O seu objectivo é fazer com que o nível de emissões de gases poluentes esteja dentro dos padrões exigidos, sem comprometer o desempenho e conforto da condução.
O projecto escolhido teve como base o uso da tecnologia das redes neuronais artificiais (RNA) para controlo de emissões.
Relativamente à implementação, foram consideradas diferentes estruturas de controlo. A análise exposta neste trabalho trata de perceber em que medida é que a introdução das redes neuronais melhora o controlo de um processo.
Os testes de desempenho são aplicados no controlo de um motor diesel com VGT e EGR, sendo realizadas através do software MATLAB. As simulações efectuadas têm diferentes configurações de modo a tirar conclusões o mais gerais possível.
This dissertation reports the project developed in the thesis/dissertation course during the 2nd year of the Master of Electrical and Computer Engineering in the field of Automation and Systems, Department of Electrical Engineering (DEE) of the Instituto Superior de Engenharia do Porto (ISEP). With the evolution of automotive systems, an electronic system became responsible for all the engine operations, yielding the electronic engine management system. Its goal is to comply with the required standards for the level of greenhouse gas emissions, without compromising the car performance and driving comfort. The chosen project was based on the use of the technology of artificial neural networks for the control of emissions. In what concerns the implementation, it was considered different control structures. The proposed idea in this work is the introduction of artificial neural networks (ANN) to improve the process control. The performance tests are conducted through the use of MATLAB software and in the control of a diesel engine with VGT and EGR. The simulations have different settings in order to get the most general conclusions.
This dissertation reports the project developed in the thesis/dissertation course during the 2nd year of the Master of Electrical and Computer Engineering in the field of Automation and Systems, Department of Electrical Engineering (DEE) of the Instituto Superior de Engenharia do Porto (ISEP). With the evolution of automotive systems, an electronic system became responsible for all the engine operations, yielding the electronic engine management system. Its goal is to comply with the required standards for the level of greenhouse gas emissions, without compromising the car performance and driving comfort. The chosen project was based on the use of the technology of artificial neural networks for the control of emissions. In what concerns the implementation, it was considered different control structures. The proposed idea in this work is the introduction of artificial neural networks (ANN) to improve the process control. The performance tests are conducted through the use of MATLAB software and in the control of a diesel engine with VGT and EGR. The simulations have different settings in order to get the most general conclusions.
Description
Keywords
RNA PID Controlo Motor diesel VGT EGR Emissões Control Diesel engine Emissions