Optimized Inter-Turn Short Circuit Fault Diagnosis for Induction Motors using Neural Networks with LeLeRU

Abstract

Induction Motor (IM) restoration costs and downtime can be decreased by early Inter-turn short circuit fault (ISCF) detection. Due to the controller’s innate desire to generate an adjusted set of currents actually below fault conditions, fault detection of electric motors driven by an inverter with a model predictive control (MPC) algorithm becomes more difficult in inverter-driven applications. We suggest a novel actuation method in this contribution using the switching sequences produced by the Finite Control Set Model Predictive Controller (FCS-MPC) for ISCF of IM. based on diagnostics from neural networks (NN). Hence, no extra sensors or equipment are required for fault detection. This paper proposes a novel procedure for ISCF fault location of IM based on Neural Networks with Learnable Leaky ReLU (LeLeLU) function.