A fault diagnosis method for inter-turn short circuit in permanent magnet synchronous motors based on optimized neural network architecture

Inter-Turn Short Circuit (ITSC) faults are a typical and severe fault type in permanent magnet synchronous motors (PMSMs). Timely and accurate diagnosis of such faults significantly enhances motor safety and reliability. To this end, this study proposes a PMSM fault diagnosis method based on an optimized neural network architecture for precise ITSC fault detection. First, an improved gram angular field is employed for feature enhancement, transforming current signals into image data with enriched feature information. Then, a hybrid model integrating an enhanced convolutional neural network and a bidirectional gated recurrent unit is constructed to deeply explore the spatiotemporal characteristics of the data. Furthermore, an improved red-billed blue magpie optimization algorithm is introduced to adaptively optimize the key hyperparameters of the neural network, further enhancing model performance. The results demonstrate that the proposed method effectively improves fault diagnosis accuracy.