Engenharia Química

URI Permanente desta comunidade

http://repositorio.ufes.br/handle/10/17472
Programa de Pós-Graduação em Engenharia Química

Centro: CCAE
Telefone: (28) 3552 8719
URL do programa: http://www.engenhariaquimica.alegre.ufes.br/pos-graduacao/PPEQ

Navegar

Navegando Engenharia Química por Assunto "Adsorption"

Agora exibindo 1 - 1 de 1
  • Nenhuma Miniatura disponível
    Item
    Estimação de parâmetros e estados via filtros Bayesianos em um processo de adsorção de SO_4^(-2) em resinas de troca iônica.
    (Universidade Federal do Espírito Santo, 2018-02-23) Carvalho, Rhynara Niandra Silva de; Dutra, Julio Cesar Sampaio; Guimarães, Damaris; Silva, Wellington Betencurte da; Pinheiro, Christiano Jorge Gomes; Costa, José Mir Justino da
    In the adsorption processes in fixed bed columns, which is very applied for the treatment of effluents, the measurements obtained at the end of the column allow to identify the efficiency of the process, but are accompanied by noise from measuring equipment. Since such systems are considered to be dynamic, therefore, the realtime monitoring of this process becomes a major challenge for the industry in order to providing a more accurate analysis of the process. Thus, the present work demonstrates the efficiency of the particle filter known as Sampling Importance Resampling (SIR) as a fast and robust tool for estimation and monitoring in a problem of sulfate ions (SO4 -2 ) removal in a fixed bed adsorption column. The fixed bed model and the mass transfer rate model were used to represent the whole process and experimental measures were used to validate the methodology used. Particle filter performance was evaluated through Root Mean Square Error (RMS) and Mean Absolute Deviation (MAD) analysis, as well as computational time verification. It was observed that the generated particles did not show degeneration by means of the verification of effective sample size (Neff) and the SIR filter presented better performance in the estimation when a higher value was used for the number of particles (Np) and for the deviation of the model of evolution (sm). In addition, flow estimation was of great importance for the filter to be able to make reliable estimates. In the analysis of the isotherms, the choice did not influence the estimation of the states. It was observed satisfactory results were observed and with remarkable applicability, once the use of the SIR filter has not yet been used to this type of process in the literature.