【摘要】 毛纽扣连接器是一种新型的垂直互连机电元器件，一般主要由毛纽扣接触件、绝缘安装板以及印制板等组成。毛纽扣接触件是该类电连接器承担导通和分断负载的核心零件，通过构建毛纽扣接触件数字样机模型，基于试验及检测对毛纽扣接触件性能退化规律进行研究，建立了基于Wiener随机过程的退化模型，在此基础上应用LSTM深度学习神经网络模型，实现毛纽扣接触件压合力、回弹和电阻随时间退化的快速计算。对毛纽扣接触件125℃使用条件下的中位寿命开展优化设计，最终获得最优材料特性参数、工艺制造参数集合及可靠性模型。可靠性优化后纽扣接触件在125℃的中位寿命为77.5h，满足使用寿命的要求。更多还原

【Abstract】 The fuzz button connector represents a novel type of vertical interconnection electromechanical component,which mainly consists of a fuzz button contact, an insulating mounting plate, and a printed circuit board(PCB). The fuzz button contact is the core component of this type of electrical connector that is responsible for conducting and breaking loads.By constructing a digital prototype model of the fuzz button contact, and based on experiments and tests, the research was conducted on the performance degradation pattern of the fuzz button contact, and a degradation model was established based on the Wiener stochastic processes. On this basis, an LSTM deep learning neural network model was applied to achieve rapid calculation of the degradation of pressing force, rebound, and resistance of the fuzz button contact over time. The optimal design of the median life of the wool button contact at 125℃was carried out, and the optimal material characteristic parameters, process manufacturing parameters and reliability model were finally obtained. After reliability optimization, the median life of the button contact at 125℃ was 77.5 h, which could meet the service life requirements.更多还原