Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

Controlo de sistemas não lineares com redes neuronais

Authors: Antunes, José Paulo Brandão;

Controlo de sistemas não lineares com redes neuronais

Abstract

O objetivo desta dissertação consiste no desenvolvimento de um controlador MPC baseado em redes neuronais artificiais para o controlo de sistemas não lineares. Para implementação de modelos de redes neuronais, que servem como identificador do modelo do sistema, é utilizada uma ferramenta com base na biblioteca TensorFlow que utiliza um método de treino de redes com base em SGD. Este método apresenta uma melhor estabilidade e capacidade de convergência quando aplicado em modelos com elevado número de parâmetros a identificar. Com este modelo é utilizado um método para síntese de controladores. É utilizado um método indireto baseado em MPC. Devido à natureza não linear do modelo é utilizado um método iterativo para encontrar a atuação ótima do controlador, baseado num algoritmo de otimização BFGS.

The main objective of this dissertation is the development of a MPC controller based in artificial neural networks in order to control nonlinear systems. For the implementation of neural network models, which serve as system model identifier, a tool based on the TensorFlow library is used that uses a network training method based on SGD. A method that presents better stability and convergence when applied in models with a high number of parameters to be identified. With this model a method for synthesizing controllers is used. An indirect method based on MPC is used. Due to the nonlinear nature of the model, an iterative method is used to find the optimum performance of the controller, based on an optimization algorithm BFGS.

Mestrado em Engenharia Eletrónica e Telecomunicações

Country
Portugal
Related Organizations
Keywords

MATLAB, SIMULINK, Tensorflow, Model-based Predictive Control, Controlo de sistemas, Identificação de sistemas, Redes neuronais, PLECS

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Upload OA version
Are you the author of this publication? Upload your Open Access version to Zenodo!
It’s fast and easy, just two clicks!