Short switch fault diagnosis method for power converter using a model‐based approach in switched reluctance motor drives

Abstract

Switched reluctance motor (SRM) is a competitive solution for aircraft and automotive applications due to its fault-tolerant capabilities and robust configuration. Power converter is a key component in SRM drives, but it prone to faults. In this study, a new and fast diagnosis method for short-circuit fault of the SRM converter is proposed based on the mixed logical dynamic (MLD) model and residual generation. Effects of short-circuit fault on SRM drives are analysed in detail. The MLD model of an asymmetrical half-bridge converter of SRM drives is presented to evaluate the currents of the power converter. The short-circuit faulty switch is identified by analysing the system residual, achieved by comparing the outputs of the MLD model and the sensor measurements. In contrast to other methods, the presented method requires no additional hardware but only measured phase currents and command signals, which are available in the control system. Simulation results confirm the effectiveness of the proposed diagnosis method. It is shown that such diagnosis method can identify the faulty switch of the power converter in a few milliseconds, which is important to avoid catastrophic consequences. It is fast, reliable and easy for industrial applications due the low-computational requirement.