Research On IGBT Reliability Analysis Method Based On Intelligent Information Processing Technology

In the past decades,wind power has made great progress.At present,wind power generation is the most probably new energy sources to replace traditional energy sources.According to the European Wind Energy Association report,A wind turbine works 3 to 6 months to obtain the energy equivalent to the manufacturing,installation,and the energy required to run out of the wind turbine.With the rapid development of new energy power generation,the IGBT module,which plays a very important role in the generation of new energy,has been paid more and more attention by people.After more than thirty years of development,the insulated gate bipolar transistor(IGBT)is widely used in large power scenarios.By exploring active and passive solutions,mechanical structure,packaging design and control strategy to study the reliability of IGBT.Even in the complex and harsh working environment,IGBT also requires a high reliability.According to industrial survey research shows that in the power conversion system,semiconductor and welding lead to the failure of the overall failure of 34%,in all the failure problem,the problem is the highest proportion.So it is very necessary to study the failure mechanism of IGBT to improve the reliability of IGBT.The main work of this paper is as follows:1.Before the analysis of the reliability of IGBT,need for a thorough understanding of the performance models of the IGBT module.In MATLAB to establish the IGBT simulation circuit,including the dynamic characteristics of the IGBT,the static characteristics of the IGBT and the simulation of the thermal characteristics of the IGBT,analysis of the impact of IGBT parameters on the IGBT characteristics.2.With the development of IGBT,the traditional IGBT reliability is very high.Experimental and time cost is too high,if the various parameters of IGBT are measured by using the normal IGBT to the failure mode.So through the accelerated life test to shorten the experimental cycle,reduce the time cost.This paper analyzes the feasibility of accelerated life test,the accelerated life test released by NASA,the analysis of the failure of what data changes significantly.3.Using intelligent information processing technology random forest algorithm,establishing of IGBT package temperature monitoring model..Using the electrical parameters of the accelerated life test data released by NASA and the input of the heat sink temperature as the input of the algorithm,the package temperature of the module is used as the output of the algorithm..The model is used as a state monitoring model of IGBT packaging temperature.Finally,the fitting degree of the model output and the actual package temperature are obtained by the experimental results..The R value is in the acceptable range,which shows that the algorithm is better fit for the packaging temperature.4.Based on the IGBT data of large quantities,the failure rate of IGBT in a certain time period is studied..The failure rate of IGBT in a certain time period is fitted by the random forest algorithm and the improved random forest algorithm,and the failure rate of the improved random forest algorithm is better compared with the random forest algorithm.