Thermal-Electrical Fusion for Real-Time Condition Monitoring of IGBT Modules in Transportation Systems

The operational reliability of Insulated Gate Bipolar Transistor (IGBT) modules in demanding transportation applications, such as traction systems, is critically challenged by solder layer and bond wire failures under cyclic thermal stress. To address this, this paper proposes a novel health monitoring framework that innovatively synergizes micro-scale spatial thermal analysis with microsecond electrical dynamics inversion. The method requires only non-invasive temperature measurements on the module baseplate and utilizes standard electrical signals (load current, duty cycle, switching frequency, DC-link voltage) readily available from the converter's controller, enabling simultaneous diagnosis without dedicated voltage or high-bandwidth current sensors. First, a non-invasive assessment of solder layer fatigue is achieved by correlating the normalized thermal gradient (∇TP) on the baseplate with the underlying thermal impedance (ZJC). Second, for bond wire aging, a cost-effective inversion algorithm estimates the on-state voltage (Vce,on) by calculating the total power loss from temperature, isolating the conduction loss (Pcond) with the aid of a Foster-model-based junction temperature (TJ) estimate, and finally computing Vce,on at a unique current inflection point (IC,inf) to nullify TJ dependency. Third, the health states from both failure modes are fused for comprehensive condition evaluation. Experimental validation confirms the method's accuracy in tracking both degradation modes. This work provides a practical and economical solution for online IGBT condition monitoring, enhancing the predictive maintenance and operational safety of transportation electrification systems.