论文部分内容阅读
基于神经网络技术和试验数据,构建了Sn-Cu焊接腐蚀性预测模型,并进行了试验验证和生产线应用。结果表明,该神经网络预测模型各输出参数的相对训练误差均小于3.5%、相对预测误差均小于4%;神经网络预测的焊接参数(卤化物含量0.06%、预热温度110-130-155℃、走板速度(1.6±0.1)m/min、焊接温度(263±2)℃和焊接时间4 s),可以满足无铜镜腐蚀、无铜板腐蚀和耐盐雾腐蚀性能好的现场技术要求,实现耐蚀性优异的Sn-Cu焊接,焊点的质量损失率降低了93.1%。
Based on neural network technology and experimental data, a corrosion prediction model of Sn-Cu welding was established, and the experimental verification and application of the production line were carried out. The results show that the relative training errors of all the output parameters of the neural network prediction model are less than 3.5% and the relative prediction errors are both less than 4%. The neural network predicts the welding parameters (halide content 0.06%, preheating temperature 110-130-155 ℃ (1.6 ± 0.1) m / min, welding temperature (263 ± 2) ℃ and welding time 4 s), which can meet the technical requirements of the field without copper mirror corrosion, no corrosion of copper plate and salt spray corrosion resistance. Achieve excellent corrosion resistance of Sn-Cu welding, solder joint mass loss rate decreased by 93.1%.