이상빈 연구실

Since electric machines have wide applicability, they can be exposed to numerous defects such as eccentricity, winding faults, lamination faults, broken bars, magnet defects, vibrations, and more. As time goes on, defects in rotating machines tend to become more severe and induce critical issues throughout the system. Therefore, condition monitoring is essential for early detection, minimizing downtime, and preventing catastrophic failures. Considering challenges identified by various manufacturers and academia, our laboratory aims to propose novel or improved detection methods with enhanced sensitivity, cost-effectiveness, and performance for various types of machine failures. Mechanical defects in PMSMs, such as load imbalance and misalignment, can lead to increased torque ripple, vibration, and performance degradation. If these issues are not detected early, they can accelerate the wear and damage of system components. While existing research has focused on detecting such faults, there is a lack of studies on developing composite control methods that both detect and control load imbalance issues. Transformers commonly experience faults like inter-laminar insulation failure, multi-point grounding faults, and transient overvoltage stress, leading to overheating and insulation breakdown. Traditional diagnostic methods such as Frequency Response Analysis (FRA), Dissolved Gas Analysis (DGA), core loss testing, and thermal imaging require offline testing or physical access, making them costly and challenging. To improve reliability, our lab works toward inventing newer approaches using existing power system components, like power converters, to inject test signals for remote and automated fault detection, enabling early diagnosis and reducing maintenance costs.