Contact characteristics and reliability assessment for high-speed electrical connector based on artificial neural network surrogate modeling

Electrical connector is an important key component in the electronic information industry, its performance and rel iability have an important effect on the performance of the entire system. In this paper, considering the influence of uncertainties from structural parameters and working load, a reliability estimation method is proposed based on physical of failure modelling and Monte Carlo simulation (MCS) method. Firstly, considering the randomness of the structural parameters and temperature, the mechanical properties of the electrical connector are studied by finite element analysis method. Further, the contact resistance is calculated by empirical theoretical model. Then, the failure modes of the electrical connector are analysed, and physical model of failure is formulated based on backpropagation neural network (BPNN). On this basis, the reliability of electrical connector is analysed by combining surrogate model and MCS method. Reliability analysis of a high-speed electrical connector is presented to validate the proposed method. The reliability evaluation results show that this type of electrical connector meets the reliability requirements.