Machine learning-based prediction of component self-alignment in vapour phase and infrared soldering

摘要: This paper aims to investigate the self-alignment of 0603 size (1.5 × 0.75 mm) chip resistors, which were soldered by infrared or vapour phase soldering. The results used for establishing an artificial neural network predicting component movement during soldering.,The components onto FR4 testboard, was designed facilitate measuring position both prior and after A semi-automatic placement machine misplaced intentionally, ability determined soldering techniques An network-based prediction method established, is able predict resistors as a function misplacement showed that can self-align from farer distances using method, even relative 50 per cent parallel shorter side component. Components only 30 method. established with approximately 10-20 mean absolute error.,It proven more stable component’s point view. Furthermore, learning-based predictors be applied in field reflow technology, networks appropriately low error.