Optimization of the Reliability of Ball Grid Array Solder Joints Under Random Vibration Conditions Using an Improved Back Propagation Neural Network-Based Particle Swarm Algorithm

Abstract Electronic products are inevitably subjected to vibration loads during shipment and use. As a mainstream packaging technology, it is of great significance to improve the reliability of ball grid array (BGA) under random vibration conditions. The aim of this paper is to improve the reliability of BGA solder joints by optimizing the structural parameters of the BGA package to reduce the stress value of the solder joints. In order to improve the design efficiency and accuracy of the computational model, the power spectral density (PSD) acceleration power spectral density is used to simulate the actual working conditions of BGA solder joints. And finite element simulation is performed in WORKBENCH software. In this paper, an intelligent algorithm combining response surface method and BP (back propagation) neural network-particle swarm algorithm is proposed to improve the accuracy of the system. It is able to find the combination of the minimum equivalent stress value of the welded joint using limited data, which solves the problem of insufficient generalization ability of the conventional method. The combination obtained by this method is verified by simulation to obtain a minimum stress value of only 0.44077 MPa, which is 21.80% lower than that of the unoptimized BGA structure. The optimized lifetime is approximately 7.316 times that of the pre-optimized lifetime. This paper provides a new idea for the optimization of structural parameters of BGA packages.