基于BP神经网络的激光叠焊焊接接头熔深预测研究

分析了激光功率、焊接速度和离焦量3个焊接参数的变化对激光叠焊焊接接头熔宽和熔深的影响,证明了熔宽和熔深的变化规律具有一致性。利用焊接参数和超声波检测信号建立了BP神经网络模型,模型验证结果表明,熔深预测的最大偏差不超过0.1 mm,最大相对误差为3%。所建立的BP神经网络预测模型满足实际应用中对激光叠焊焊接接头熔深测量要求。

Research on Laser Lap Weld Depth Prediction Based on BP Neural Network

The influence of laser power,welding speed and defocusing distance on laser lap weld width and depth was analyzed.It is proved that the changing patterns of weld width and depth are consistent.A BP neural network prediction model for laser weld depth is established by the welding parameters and ultrasonic testing signals.The verification results of the model show that the maximum deviation of the weld depth prediction is less than 0.1 mm,and the maximum relative error is 3%,which meets the measuring requirements of laser weld depth in practical applications.