Prediction of Laser Cutting Quality for QFN Package by using Neural Network

This paper reports the predictions of laser cutting QFN (quad flat non-lead) packages by using Levenberg-Marquardt backpropagation algorithm of neural network. A mathematical model via neural network was proposed for predicting the laser six cutting quality. A 5times5 QFN package was cutting by using a diode pumped solid state laser system (DPSSL) in this paper. From the predicted results, six average errors of cutting qualities are 0.6%, 1.24%, 3.21%, 2.44%, 5.08% and 13.73%. The results may give guides in the predictions of cutting QFN packages and is expected to be useful for laser applications in other industry fields