A Sensor-less IGBT On-State Voltage Estimation Method Using Inverter Control Variables

The on-state voltage is a commonly used health indicator for monitoring the health status of IGBT modules. However, deriving the on-state voltage through measurement-based methods requires additional measurement circuits, which increases both the cost and volume. Existing sensor-less methods, meanwhile, depend largely on training datasets to extract degradation features specific to particular converters, limiting their wider applications. To address these challenges, a sensor-less method uses variables collected for control purposes to estimate the on-state voltage and current is proposed. The proposed method eliminates the need for a training dataset. Additionally, the impact of noise in control variables is mitigated by averaging a series of samples collected under similar operating conditions to improve estimation accuracy. The effectiveness of the proposed method has been validated through hardware-in-the-loop experiments conducted across three case studies, each representing different degradation stages, power levels, and power factors. The estimation errors at full load are close to 1% of the initial on-state resistance.