Incipient detection of stator inter‐turn short‐circuit faults in a Doubly‐Fed Induction Generator using deep learning

Wind turbines are increasingly expanding worldwide and Doubly-Fed Induction Generator (DFIG) is a key component of most them. Stator winding fault major in this equipment its incipient detection vital importance. However, there paucity research field. In study, novel machine learning-based method proposed for inter-turn short-circuit (ITF) the DFIG stator based on current signals stator. The makes use state-of-the-art deep learning methods along with conventional signal processing tools general techniques. More specifically, problem regarded as multi-class classification Long Short-Term Memory network, which more appropriate time-series data utilised inference. Furthermore, variant celebrated Empirical mode Decomposition analysis tool used to extract some well-known statistical features among informative ones selected using new feature selection method. Our tests experimental steady-state conditions show that can accurately detect ITF at initial stage when only one turn shorted. Moreover, performance considerably higher than variety methods.