Fechar
Metadados

@InProceedings{EchevarrķaVelBecSilSan:2012:ACAC,
               author = "Echevarr{\'{\i}}a, L{\'{\i}}dice Camps and Velho, Haroldo 
                         Fraga Campos and Becceneri, Jos{\'e} Carlos and Silva Neto, 
                         Antonio J. and Santiago, Orestes Llanes",
          affiliation = "ISPJAE, Cuba. and {Instituto Nacional de Pesquisas Espaciais 
                         (INPE)} and {Instituto Nacional de Pesquisas Espaciais (INPE)} and 
                         IPRJ-UERJ, Brazil. and ISPJAE, Cuba.",
                title = "The Fault Diagnosis Inverse Problem: ACO and ACO-d",
            booktitle = "Proceedings...",
                 year = "2012",
         organization = "International Symposium on Uncertainty Quantification and 
                         Stochastic Modeling, 1.",
                 note = "{Setores de Atividade: Pesquisa e desenvolvimento 
                         cient{\'{\i}}fico.}",
             keywords = "Ant Colony Optimization, fault diagnosis, inverse problem, 
                         processing time, robustness, structural detectability, structural 
                         separability.",
             abstract = "The automatic early detection, isolation, and localization of 
                         faults is of high interest in industrial systems, for improving 
                         reliability and safety. This process is characterized as fault 
                         diagnosis (FDI) and some problems related with robustness to 
                         external disturbances, sensitivity to incipient faults and 
                         processing time are still considered as limitations for many of 
                         the current FDI methods. This work is focused on the formulation 
                         of the fault diagnosis by an inverse problem methodology. The FDI 
                         problem is formulated as an optimization problem and takes results 
                         from the diagnosis area for acquiring prior information. The 
                         optimization problem is solved by the stochastic algorithm Ant 
                         Colony Optimization (ACO) and its modified version fuzzy-ACO 
                         (ACO-d). The proposed approach is tested using simulated data of 
                         the Inverted-Pendulum system which is recognized as a benchmarck 
                         for control and diagnosis. With the purpose of analyzing the 
                         advantages of such approach, some experiments, with data corrupted 
                         with noise, are considered. The influence of ACO parameters are 
                         also taken in consideration. The results obtained show the 
                         suitability of the approach and also indicate that the parameters 
                         values allowing a greater exploration of the search space yields a 
                         better diagnosis. The ACO-d algorithm enables better diagnosis 
                         than ACO.",
  conference-location = "S{\~a}o Sebasti{\~a}o, SP",
      conference-year = "Feb. 26th to Mar. 2nd",
                 issn = "2238-1007",
                label = "lattes: 1840336983035010 3 
                         Echevarr{\'{\i}}aVelBecSilSan:2012:ACFuAC",
             language = "pt",
           targetfile = "56-Lidice.pdf",
        urlaccessdate = "16 jan. 2021"
}


Fechar