Prediction of welding properties of solder joints based on machine learning

Abstract

To greatly improve the design efficiency and the prediction of the welding properties of on-site solder joints in the welding workshop, an intelligent calculation algorithm based on CATIA. API for secondary development of the solder joint level is designed. This algorithm, based on different interference collision radiation radii, calculates the correlation accuracy between different solder joints and BIW products, and extracts the BIW product information data related to the solder joints, to realize the automatic generation technology of digital and intelligent welding attribute BOM. On this basis, we use Python to write the linear regression algorithm in machine learning to train the welding attribute parameters of the solder joints.