Cluster Analysis And Modeling Study On Performance Degeneration Of Tantalum Electrolyte Capacitor

Solid tantalum capacitor with the characteristics of high stability and reliability are widely used in all kinds of fields.The performance of tantalum capacitor directly influences the safe running of the whole system.Therefore,the performance degradation law,degradation mechanism,degradation model and degradation assessment are of great significance for running strategies and system stability of system.Based on the study on the test method and failure mechanism of degradation performance of tantalum capacitor,the cluster analysis model and the degradation model based on single parameter and dual parameter are established,and a performance degradation evaluation method is determined to finish reliability evaluation.The main work of this paper is as follows:Firstly,the research background and the significance of this paper are explained.The research status on degradation performance,failure mechanism,self-healing characteristics and performance degradation modeling of tantalum capacitor are introduced.Then the research contents of this paper are presented,mainly including the development of test platform,the acquirement of degradation data,the analysis of failure mode and mechanism,the determination of cluster analysis method of degradation performance,the establishment of the degradation models and the performance degradation evaluation methods based on single parameter and dual parameters.Secondly,the structure and main degradation performance parameters of tantalum capacitor are analyzed.The accelerated degradation test method is designed.With the variable interval measurement,the constant stress as the accelerated test type,temperature and humidity as stress parameters,a degradation test platform is built and the degradation data is obtained.Then the failure mode and the main degradation mechanism of tantalum capacitor are analyzed,which provide a basis for cluster analysis and degradation modeling of performance degradation.Then,in view of the fact that the change of the degradation parameters of tantalum capacitor with temperature and humidity is not a single trend change,the ordered clustering method is introduced.The natural property and the change rate α(k)of the slope β(k)of the error function δ[p(m,k)] are added and the determination method of the optimization clustering number is put forward to.The Davies-Bouldin Index is proposed as evaluation method of ordered clustering result.The validation of the evaluation method is verified by the degradation data with other stresses.Finally,a single-parameter degradation fitting method for tantalum capacitor based on the Least Squares method is proposed.The coefficients of the degradation model at stress level of 95 °C and 70% RH are determined by use of the determination coefficient and the root mean square error,which is verified with the degradation trend at 85 °C and 85% RH.After the key degradation parameters of tantalum capacitor are determined through the life distribution test and the pseudo life prediction,an establishment method of accelerated degradation model based on Wiener process with random effects is proposed.A binary accelerated degradation model based on the Copula function is determined.The performance evaluation results show that the binary accelerated degradation model can be performed on life prediction of tantalum capacitors...