A diagnosis approach for open-circuit faults in NPC three-level energy storage converter

Condition monitoring and fault diagnostic techniques are important to enhance the reliability of energy storage systems. This paper proposes a switching-signal-based open-circuit fault diagnosis approach for the IGBT switches of NPC three-level energy storage converter, enabling the rapid detection and identification of multi-switch open-circuit faults. In response to the complexity of fault behaviors and the difficulty in distinguishing fault characteristics in multilevel converter, a state model under the fault mode of NPC three-level energy storage converter is developed. This model can effectively track the abnormal transitions of the circuit's operating state before and after single and multi-switch open-circuit faults, and consequently obtain the diagnostic variables that reflect these abnormal state transitions. Subsequently, the state variables related to the abnormal transitions are optimized to derive the minimum feature set necessary to represent various fault modes. Through detailed analysis of the logical relationships between characteristic set variables and switching signals, the open-circuit fault signals for each IGBT are derived. Simulation results demonstrate the proposed approach can swiftly locate multiple faulty switches simultaneously, without requiring complex computations and possessing capabilities.