Research On Accelerated Test Optimum Design And Data Statistical Analysis Of Electrical Connectors Based On Stochastic Process

As a basic electromechanical component to transmit signals and electrical energy in model equipment,electrical connectors ploy an important role in connection and separation between the equipment and the ground,among units and among systems,but it is a weak link on the entire model.Therefore,it is crucial to accurately evaluate and extension the life of model equipment in reliability assessment of electrical connector.During the model equipment's storage,electrical connector are mainly affected by temperature.However,mixed-effects model do not reflect the randomness of performance in describing the degradation of performance,which will cause a large deviation in the final assessment.Therefore,in this paper,the Y11P-1419 electrical connector is used as the analysis object to describe its performance degradation based on stochastic process modeling,determine the final plan after the optimization and simulation,evaluate the reliability of storage.These provide the theoretical basis for reliability assessment.In Chapter 1,the research background and the significance of the dissertation are clarified.In this chapter,the domestic and foreign research status are analyzed including the accelerated degradation model,the optimal design of test plan,the statistical analysis of the test data and the reliability of the electrical connector.On this basis,the research approach and the main research content are proposed.Chapter 2 analyzes the generation and gradual increase of oxidation corrosion products in electrical connector's contact.According to the random encounter between oxide corrosion and contact spots in the contact surface,the nonlinear Wiener Process of stochastic process describing the performance degradation,the life distribution model of the reliability statistical model are established.In Chapter 3,according to the stochastic process model,designed variables and some constraints,a mathematical model of experiment optimization is established.Based on this,a traditional test optimized plan and two-stress level unequal distribution test optimized plan are discussed by comparing them from perspectives of estimated accuracy and robustness.Then,combined with the actual test environment,the final compromise optimization test plan is established.In Chapter 4,by the mean and variation coefficient of the asymptotic variance of the estimated lifetime under the storage environment as assessment indicators,the estimated accuracy and the robustness of the test plan are evaluated based on Monte Carlo simulated.Next,the correctness of the test plan optimization design theory based on stochastic process is verified.In Chapter 5,based on the maximum likelihood estimation theory,the parameters of the nonlinear Wiener Process model are estimated,and then the interval estimates of drift coefficient and diffusion coefficient are obtained based on regression analysis.On this basis,the reliability characteristic value is estimated through the reliability statistical model.By the variance analysis and regression analysis,the acceleration equation is tested.By using the correlation coefficient theory of Person,the linear correlation between simulated data produced by the performance degradation model and experimental test data was analyzed,and the normality test of the performance variation verify the correctness of the performance degradation model by the result of correlation.Finally,the research work of this dissertation is summarized,and the future research is forecasted in Chapter 6.