Prediction on health of welded point based on gray model

With the wider use of FPGA with BGA encapsulation in the aerospace and weapon system, the health status of welded point is attracting more and more attention. In this paper, the gray system theory is used to solve problems such as lack of data points, no obvious regularity, difficult to predict, etc., while establishing failure model for welded point. A failure model of welded point is established using the data measured during fault diagnosis as original data. The prediction about the later health status of welded point is made, and the comparison of predicted and measured results reveals that the prediction accuracy is in level 1. The experimental results show that the prediction model for health status of welded point based on gray system theory can alert before the failure of welded point, and overcome the problems of long cycle, being difficult to predict failure of welded point and instant failure after being found.