Fault Diagnosis of Inter-Turn Short Circuits in Doubly-Fed Induction Generators Based on External Magnetic Field and Vibration Signal Fusion

This manuscript introduces a novel diagnostic technique for identifying inter-turn short circuits in doubly-fed induction generators (DFIGs) by integrating multi-channel external magnetic field (MFL) and vibration signals. The approach begins with cepstral pre-whitening to mitigate noise and improve the identification of fault-related characteristics. Following this, a correlation analysis is conducted between vibration signal and external magnetic field signals at various locations on the generator. The correlation coefficients obtained are then employed to calculate the vibration-MFL signal cooperative gain, which measures the interaction between these signals. To enhance fault detection, a vibration-MFL signal cooperative index is developed by combining the cooperative gain with the correlation coefficients, providing a comprehensive metric for diagnosing inter-turn short circuit faults. This method's effectiveness is validated on a 100 kW DFIG wind turbine simulation platform for inter-turn short circuit faults. The results confirm that this technique can reliably detect early-stage inter-turn short circuits, thereby enhancing the operational reliability and safety of DFIGs in wind energy systems.