Research On Junction Temperature Estimation Strategy Of IGBT Module For Electric Vehicle Considering Multiple Working Conditions And Boundary Conditions

In recent years,high-power insulated gate bipolar transistor(IGBT)has been widely used in new energy power generation,smart grid,electric vehicles,national defense and security.As the key device of power electronic system,its reliability has attracted more and more attention.IGBT junction temperature information and junction temperature fluctuation are directly related to module temperature protection,device life prediction and analysis,converter thermal management.Accurate acquisition of IGBT junction temperature information is an important way to improve the reliability for chip level,module level and system level.From the view point of engineering application,this paper takes the IGBT power module for electric vehicle as the research object and the online prediction for chip junction temperature as the research content.Based on the finite element analysis,the temperature field distribution in power module for electric vehicle under different thermal boundary conditions is explored,which simplifies the thermal network modeling and model parameter identification,and then puts forward a "fast and accurate" junction temperature estimation method suitable for the module.Through the towing platform and quasi correction for power loss model under the locked rotor condition,the accuracy of the proposed scheme in the multi working conditions and complex thermal boundary conditions is verified.In this paper,some problems that need to be paid attention to in the online junction temperature estimation for power modules in electric vehicles is analyzed,such as accuracy requirements,cost issues,module invasiveness,etc.,after full consideration,we intend to realize the accurate junction temperature detection based on the electro-thermal coupled thermal network.In order to ensure the junction temperature estimation accuracy,the switching loss model is obtained based on double pulse test,and then the loss distribution for inverter is obtained.As the input of thermal network,it directly affects the subsequent junction temperature estimation.In view of the compact chip layout and unique direct liquid cooling structure of the selected IGBT module,the maximum junction temperature monitoring point for the module is determined through experiments and finite element analysis.At the same time,the thermal network from chip to coolant can be directly established due to its direct substrate structure.In the ANSYS/icepak environment,this paper analyzes the thermal impedance between the chip junction and different reference points under different thermal boundary conditions,and puts forward the strategy that the chip junction temperature in high side is predicted by NTC thermistor,and that in low side is predicted by coolant temperature,and puts forward two schemes according to whether there is a coolant temperature detection unit.Based on the motor towing platform and junction temperature measurement system,the proposed scheme is verified in the MATLAB / Simulink environment,and the motor locked rotor condition is introduced in detail.Its principle and platform conditions provide the possibility for quasi online correction for chip power loss,and then the process of correcting power loss model for IGBT module based on locked rotor condition is proposed.Considering the various operating conditions for electric vehicles and the complex thermal boundary conditions in the motor controller,this paper simulates the mission profile for the motor power converter by giving the torque and speed,and simulates different boundary conditions by changing the coolant temperature and flow rate,so as to verify the proposed method accuracy,calculation speed and adaptability.The efficiency and practicability of the temperature estimation strategy are also proved.