作者: Marcos F.S.V. D’Angelo , Reinaldo M. Palhares , Murilo C.O. Camargos Filho , Renato D. Maia , João B. Mendes
DOI: 10.1016/J.ASOC.2016.08.040
关键词:
摘要: Graphical abstractDisplay Omitted HighlightsA new data-driven fault detection and isolation scheme is presented.The fuzzy/Bayesian approach used to indicate a possible event.The faults classification performed using immune/neural approach. This study presents data-based methodology for in dynamic systems based on change point associated with hybrid formulation pattern applied the Tennessee Eastman benchmark process. The detected when occurs signals from sensors classified into one of classes by formulation. system fuzzy set theory MetropolisHastings algorithm system, main contribution this paper representation which combines ClonALG Kohonen neural network.