A High-precision Adaptive Thermal Network Model For Monitoring Of Temperature Variations In Power Semiconductor Devices

As a core component in power converter systems,insulated gate bipolar transistors(IGBTs)are widely used in safety-critical and high-reliability systems,e.g.electrical vehicles,aircraft,wind turbines,and industrial drives.Terrible operating conditions could harm the reliability of the devices,such as high current intensity,high voltage and high switching frequency.According to the relevant data,it shows that more than 50% of power electronic systems breakdown are attributed to excessive junction temperature.Therefore,in order to improve system reliability,it is necessary to have a good understanding about the relation between junction temperature and remain useful life of IGBTs.Based on temperature characteristics,the Cauer-type thermal network model was established in PLECS to obtain the relationship between physical parameters and junction temperature.A new method is proposed to improve the performance of the temperature estimation of IGBT modules.The conventional Cauer-type thermal network model was modified from steady-state junction temperature estimation and transient temperature.What's more,based on mechanism of solder fatigue,this paper proposes an effective heat propagation path-based online adaptive thermal model for IGBT modules,where the effective heat propagation path is proposed to quantify the impact of substrate solder crack and the degree of nonuniformity of temperature distribution is established.The proposed adaptive thermal model is validated by comparing with finite element analysis(FEA)simulation results for a commercial IGBT module.In order to validate precisely,experimental platform was established.The results show that the proposed thermal network not only has great performance of junction temperature estimation but can track the thermal behavior changes of the IGBT modules caused by substrate solder cracks.