Please use this identifier to cite or link to this item:
|Type:||Artigo de evento|
|Title:||Advanced Control Strategy To A Fermentation Process To Obtain Ethanol|
Maciel Filho R.
|Abstract:||The aim of this work is to evaluate the performance of a non linear control strategy based on Artificial Neural Networks (ANN's) for an extractive alcoholic fermentation process (MIMO 3x3). The process is simulated through a validated deterministic model which takes into account the main phenomena taking place in the system. The control is build-up by a multilayered and multivariable ANN that represents the inverse dynamics of the system, which is on-line trained through an optimization routine. The optimization routine adjusts the weight of neural controller using the estimate global error of closed loop. The global error is obtained using a dynamic model of the process, to represent a prediction for the next sampling time. This process model is on-line trained with the process data. The obtained results have shown the efficiency of control strategy in multivariable system and the potential of proposed on-line learning.|
|Appears in Collections:||Unicamp - Artigos e Outros Documentos|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.