基于电极位移信号的电阻点焊质量预测模型

设计了电阻点焊过程动态参数监测系统，采集焊接电压、电流以及电极位移信号，建立以交流电阻点焊过程中的周波参数为输入空间，以试样的抗拉剪切强度为输出空间的点焊质量预测神经网络模型．为实现电阻点焊质量在线监控奠定了理论基础。试验结果证明了RBF神经网络用于这类应用场合的可行性和有效性。

Quality forecasting of resistance spot welding joints

A dynamic parameters monitoring system was extablished to collect welding voltage,welding current and electrode displacement signals during resistance spot welding and to set up neural network,in which the cycle parameters of dynamic resistance and displacement were used for inputting parameters,and the corresponding shear strength of samples are used for outputting parameters,which established the essential theoretic foundation for realizing the on line quality controlling in spot welding process.The results of the experimentats prove that it is effective for RBF neural network being used in this field used for these type occasions.