Wideband Frequency Response Analysis for Sensitive Condition Estimation of Machine's Turn Insulation Degradation Faults

Advances in power electronics-based drives have made electric machines crucial for efficient and reliable energy conversion. However, stator turn insulation failure remains a primary cause of inverter-fed motor breakdowns. Effective condition monitoring of turn insulation is vital for mitigating safety risks but is often challenging due to weak characteristic variations in the signal. This paper presents a robust approach for evaluating stator turn insulation health using a proposed grey-box optimal model. High-frequency switching oscillation currents are analyzed using fractional domain methods to enhance sensitivity in the detection of insulation degradation. Experimental results from a 0.75 kW induction motor test setup with degraded cases demonstrate the effectiveness of the proposed method. By leveraging the fractional Fourier transform, this approach enhances turn insulation state estimation sensitivity by approximately four-fold compared to the conventional transform technique at early stages.