01800nmc a2200277 4500 20260520005606.0 260224s2001 xxu ing Inglés ARTIFICIAL NEURAL NETWORKS APPLIED TO THE OPERATION OF A FLUID CATALYTIC CRACKING (FCC) UNIT SPE 69501 Dallas, Texas Society of Petroleum Engineers 2001 12/09/06 ; 17/07/01 A model, based on artificial neural networks (ANN), for a fluid catalytic cracking (FCC) unit is developed. The ANN model of the plant was developed using process and laboratory data directly from a refinery data acquisition system. And it used to predict process variables and estimate several product quality properties. The empirical model was accurate up to 98.0 % when tested on experimental data not used during training, which is an outstanding performance. The present paper shows how a neural network, three-layered perceptron, can be used effectively to simulate the actual behaviour of refining units, specifically a fluid catalytic cracking (FCC) unit. CONGTP Pérez, José R. 6164 López, Roberto 6165 Dassori, Carlos G. 6166 Ranson, Aaron 6167 Latin American and Caribbean Petroleum Engineering Conference (7th : 2001 mar. 25-28 : Buenos Aires, Argentina) 5738 López, Roberto 6165 Dassori, Carlos G. 6166 Ranson, Aaron 6167 Latin American and Caribbean Petroleum Engineering Conference (7th : 2001 mar. 25-28 : Buenos Aires, Argentina) 5738 Pérez, José R. 6164 131012 131012 https://biblioteca.iapg.org.ar/ArchivosAdjuntos/LACPEC2001/Spe69501.pdf 0 0 0 0 201934 BVIAPG BVIAPG 2026-03-03 200003067 0 200003067 2026-03-03 01:41:09 2026-03-03 ARCHIVO