Neurofuzzy controller in automated climatization for poultry houses

This work presents a Neurofuzzy Controller (NFC) based on the Takagi-Sugeno fuzzy model for use in non-linear plants when complexity does not allow setting a good mathematical model for the system. A comparisons was made between a neurofuzzy controller and a neural controller based on the Multi-Layer-Perceptron Network. Conclusions were withdrawn from a prototype of a poultry house climatization system in a 1:17 laboratory scale. The results show the excellent performance of the proposed controller due to an easy incorporation of the human specialists' knowledge and the learning capacity of the artificial neural network.