Agarwal, DeepeshDeepeshAgarwalKumari, ManikaManikaKumariSrinivasan, BabjiBabjiSrinivasanRajappan, PrabhavathyPrabhavathyRajappanChinnachamy, JaishankarJaishankarChinnachamy2025-08-312025-08-312020-01-01[9781728142517]10.1109/PESGRE45664.2020.90702812-s2.0-85084286522https://d8.irins.org/handle/IITG2025/24340Electric motors are solely responsible for producing traction power in Electric Vehicles (EVs). It is imperative to conduct health monitoring of motors in order to ensure robust vehicle performance, adhere to safety requirements and avoid further catastrophic consequences leading to failure of the powertrain. Fault diagnosis and degradation analysis allow to capture the motor abnormalities at an earlier stage, so that suitable preventive measures can be adopted to limit the severity of faults. Simulation studies of motors help understand the nature of various incipient faults without expensive experimentation. Further, several incipient faulty conditions are difficult to introduce in limited time experiments. Given the advantages of simulations, we perform fault diagnosis and degradation analysis using Finite Element Analysis (FEA) based motor models. Since the detailed simulations are time-consuming, we generate surrogate current data using that obtained from simulations. We employ support vector machines for fault classification using features obtained from current data. The robustness of the proposed framework to measurement noise is also analyzed.falseDegradation Analysis | Electrical motor | Fault Diagnosis | Surrogate data | Wavelet Packet DecompositionConference PaperJanuary 202069070281cpConference Proceeding3