EPE Association: EPE 2023 - LS7a: Focus Topic 6 - Reliability and Artificial Intelligence in Power Electronics

EPE 2023 - LS7a: Focus Topic 6 - Reliability and Artificial Intelligence in Power Electronics
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2023 ECCE Europe - Conference > EPE 2023 - Topic 11: Focus Topics > EPE 2023 - LS7a: Focus Topic 6 - Reliability and Artificial Intelligence in Power Electronics

   [return to parent folder]

   A model-based approach for prognostics of power semiconductor modules
By Aleksi VULLI, Gerd SCHLOTTIG, Michal ORKISZ, Marcin FIRLA, Enea BIANDA [View]
[Download]

Abstract: We developed a model-based approach for prognostics of power semiconductor modules, including novel physics-inspired damage progression models for chip and system solder degradation, stemming from temperature variations. Particle-filter-based methodology enables estimation of the module's current health, forecast of the two concurrently progressing damage mechanisms, and finally prediction of the remaining useful life. In this publication we demonstrate the approach in the case of chip solder degradation.

   Electric Vehicle Thermal Management System Modeling with Informed Neural Networks
By Ekin Alp BICER, Pascal SCHIRMER, Peter SCHREIVOGEL, Gabriele SCHRAG [View]
[Download]

Abstract: Proper modeling of Thermal Management System (TMS) in Electric Vehicles (EVs) is crucial in terms of designing the EV components. Data-driven methods come up as an alternative to the computationally intensive high-fidelity methods or reduced order models where the accuracy is sacrificed for performance. In this paper, two informed neural network approaches are benchmarked in EV TMS modeling: Analytical Feature Engineering, where new features are generated by using the physical processes that take place within the EV, and Feature Generation via Loss Maps where loss maps of the inverter and the electric engine are used to generate a new power loss feature. Results show that accuracy increased by 1.7\% to 3.6\% depending on applied features and the neural network architecture.

   Online Threshold Voltage Monitoring at SiC Power Devices during Power Cycling Test and Possible Consequences
By Patrick HEIMLER, Mohamed ALALUSS, Christian SCHWABE, Xing LIU, Josef LUTZ, Thomas BASLER [View]
[Download]

Abstract: In this paper, the threshold voltage shift during power cycling test for discrete SiC-MOSFET devices in TO-247 package from three different manufacturers with the same blocking capability of 1200 V is investigated. The temperature dependence of the threshold voltage plays a significant role in the interpretation of the results, since a change in Vth can also influence common TSEP (temperature-sensitive electrical parameter) measurement methods. The tested devices have shown no deterioration of the solder layer, but degradation of the bond wires in the failure analysis.

EPE 2023 - LS7a: Focus Topic 6 - Reliability and Artificial Intelligence in Power Electronics
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2023 ECCE Europe - Conference > EPE 2023 - Topic 11: Focus Topics > EPE 2023 - LS7a: Focus Topic 6 - Reliability and Artificial Intelligence in Power Electronics
	[return to parent folder]

	A model-based approach for prognostics of power semiconductor modules By Aleksi VULLI, Gerd SCHLOTTIG, Michal ORKISZ, Marcin FIRLA, Enea BIANDA	[View] [Download]
Abstract: We developed a model-based approach for prognostics of power semiconductor modules, including novel physics-inspired damage progression models for chip and system solder degradation, stemming from temperature variations. Particle-filter-based methodology enables estimation of the module's current health, forecast of the two concurrently progressing damage mechanisms, and finally prediction of the remaining useful life. In this publication we demonstrate the approach in the case of chip solder degradation.

	Electric Vehicle Thermal Management System Modeling with Informed Neural Networks By Ekin Alp BICER, Pascal SCHIRMER, Peter SCHREIVOGEL, Gabriele SCHRAG	[View] [Download]
Abstract: Proper modeling of Thermal Management System (TMS) in Electric Vehicles (EVs) is crucial in terms of designing the EV components. Data-driven methods come up as an alternative to the computationally intensive high-fidelity methods or reduced order models where the accuracy is sacrificed for performance. In this paper, two informed neural network approaches are benchmarked in EV TMS modeling: Analytical Feature Engineering, where new features are generated by using the physical processes that take place within the EV, and Feature Generation via Loss Maps where loss maps of the inverter and the electric engine are used to generate a new power loss feature. Results show that accuracy increased by 1.7\% to 3.6\% depending on applied features and the neural network architecture.

	Online Threshold Voltage Monitoring at SiC Power Devices during Power Cycling Test and Possible Consequences By Patrick HEIMLER, Mohamed ALALUSS, Christian SCHWABE, Xing LIU, Josef LUTZ, Thomas BASLER	[View] [Download]
Abstract: In this paper, the threshold voltage shift during power cycling test for discrete SiC-MOSFET devices in TO-247 package from three different manufacturers with the same blocking capability of 1200 V is investigated. The temperature dependence of the threshold voltage plays a significant role in the interpretation of the results, since a change in Vth can also influence common TSEP (temperature-sensitive electrical parameter) measurement methods. The tested devices have shown no deterioration of the solder layer, but degradation of the bond wires in the failure analysis.