Condition and Health Monitoring
Simultaneously, one pivotal aspect of this exploration revolves around the proactive management of potential failures in power electronics. Condition and Health Monitoring (C&HM) emerges as a key strategy to enhance the availability and control of power electronic components, converters, and systems. Despite the increasing relevance of C&HM techniques, their integration into industrial applications remains limited. This presents an opportunity for advancements that can drive their widespread industrialization. One of the main concerns continues to be the reliability of power devices and modules as they are known to be susceptible to failures, hinging on accurate predictions of future reliability and remaining useful lifetime. In this sense, in-situ methods for estimating real-time junction temperature, particularly using Thermo-Sensitive Electrical Parameters (TSEPs) and other in-module sensors become a crucial tool for active control and health monitoring in power electronic systems. However, relying only on the physical domain faces several challenges in addressing ongoing device degradation. Thus, bridging the gap between laboratory verification and field testing becomes imperative through the exploration of special operation modes and advanced data analysis techniques.
Our R&D Activities:
Estimation of Tj based on TSEP. Fast and efficient monitoring circuits will be developed which allows the use of WBG semiconductor device internal gate resistance and ON-state voltage as TSEPs to predict temperature and health-status.
New and non-conventional methods to measure temperature and estimate degradation. FLAGCHIP will adopt a novel approach using Fibre Bragg Grating sensors and package the sensors inside the power module to monitor the temperature of the power semiconductor device chips and the degradation of the packaging.
Prediction of the health status and lifetime of the WBG power modules based on on-line in-situ monitoring results. Digital twin of Power Electronics Module with real-time WBG switch model simulation engine, supplied with real-time acquired data, the simulated switch behaviour and thermal effects will be compared with real switch ones for on-line early detection of parameter discrepancy
Combination of data-driven and physics-of-failure driven approaches: FLAGCHIP will power cycle the developed power modules and obtain a big range of data offline to reflect the aging of the power semiconductor devices and packaging with different power cycling conditions regarding the power levels and frequency.
When the power modules are operated in power converters, a data driven approach will be used to infer the health status of the power modules by measuring their health monitoring signals online such as temperature, ON-state voltage, the degradation of the packaging.
To understand the physics of the failure, TCAD models of existing or new devices in SiC and other UWBG materials will be developed to provide accurate information on how the driving and operating conditions can affect the reliability and lifetime of the project devices or the commercial devices, which will allow us to develop better and more accurate condition monitoring algorithms.
