Fault detection and isolation for a multi-cellular converter based on sliding mode observer

This paper presents a model-based approach for the fault diagnosis of three-cell power converters. The goal is to design a fault detection and isolation (FDI) scheme which diagnoses parametric and discrete faults. Parametric faults are characterized by abnormal changes in the capacitor value whereas a discrete fault affects a switching element and is characterized by an abrupt deviation of the dynamics of the system. The combined continuous and discrete diagnosers are used to detect and isolate the discrete and the parametric faults. The discrete diagnoser is based on an appropriate decomposition of the hybrid model of the three-cell converter. The continuous diagnoser is based on a set of residuals which enable a comparison between the measured and estimated values of each continuous variable. Some experiments are shown to prove the effectiveness of the proposed scheme.