Research On Lifetime Prediction Of IGBT In Suspension Chopper Based On Accelerated Life Test

As the core driving device of medium and low-speed maglev trains,the suspension chopper is the key to the reliability evaluation of the whole vehicle.The power thyristor module is the core device of the suspension chopper,so the degradation process of the module should be studied.In recent years,the research on reliability and life prediction models of semiconductor power devices has received extensive attention at home and abroad,but the existing life models are not suitable for explaining the load sequence effect of semiconductor power devices,and there are limitations in explaining the module degradation process.Aiming at this problem,which is how to explain the degradation process of the suspension chopper power module and evaluate its life,the following research is carried out:1.The life prediction model of the IGBT module is established.Based on the lumped parameter method,the electrothermal model of the IGBT module is established.A data statistical model suitable for statistic IGBT module load is established.Based on the different processing methods of damage variables,different types of fatigue damage models are analyzed and compared,and a nonlinear fatigue damage model based on life curve method and damage variable correction formula is established.Finally,a model optimization method is proposed.The optimized model can explain the loading sequence effect and better explain the damage development process without considering the interaction between loads.2.Accelerated life test of the IGBT module was carried out.The acceleration property of the failure process of IGBT modules is analyzed,and a method of constant stress accelerated life test with fixed number truncation is selected.According to the accelerated failure mechanism of IGBT modules,the accelerated life test scheme of active power cycling is selected,and the corresponding control strategy is formulated.The power cycling test platform was built,the multi-physics degradation data collection was completed,and it was verified that the failure of the bonding wire was the main failure mechanism of the power cycling test.3.The analysis of the test data and the optimization of the prediction model were completed.According to the transient thermal impedance data obtained from the experiment,the parameters of the equivalent thermal model are obtained by fitting.According to the degraded state development data obtained from the test,the parameters of the three fatigue damage models are obtained by fitting,and the goodness of fit of the three models is quantitatively analyzed.The results show that the most suitable model has good interpretive ability on the degraded state development data.Finally,the optimization of the life prediction model is completed.The results show that the model using the optimization algorithm can explain the loading sequence effect and has the corresponding damage development mechanism basis.4.The key factors affecting the life of the IGBT module of the suspension chopper are analyzed.The three factors of DC bus voltage fluctuation,ambient temperature change,and parameter drift of the module are analyzed and compared,and the key issues that need to be considered for each factor are pointed out.