Fault Diagnosis and Fault-Tolerant Control Operation of Nonisolated DC–DC Converters

This paper proposes a fault diagnosis and identification method for nonisolated dc-dc converters. The method can detect the fault, and identify the faulty switch and the switch fault type in less than one switching cycle. The main idea is to apply the derivative of the inductor current based on the modulation of the particular dc-dc converter to provide a predictive current emulator model. The measured inductor current and its corresponding predicted current are used to diagnose the switch fault. The method, implemented in a digital signal processor, is robust to common converter asymmetry such as load variations, input disturbances, etc. The proposed fault diagnosis method has been analyzed, tested, and validated for a boost converter.