论文部分内容阅读
搭建了基于旋转电弧传感法的焊接试验平台,制定焊缝成形与焊接电流波形情况的评价指标,确定以旋转电弧传感器检测的焊接电流、旋转频率、焊枪高度和焊枪旋转半径作为影响焊缝成形与焊接电流波形的关键因素。进行了焊接试验,采集训练数据,推导了焊接电流、旋转频率、焊枪高度、焊枪旋转半径与焊缝成形、焊接电流波形之间的神经网络模型。最后的验证试验表明,所构建的神经网络模型能较好地预测焊缝成形及焊接电流波形情况,能够为旋转电弧法最优焊接工艺参数的选取提供依据,以实现优异的焊接过程自动化。
The welding test platform based on the rotating arc sensing method is set up and the evaluation index of the welding seam shape and welding current waveform is established. It is determined that the welding current, the rotation frequency, the torch height and the torch rotation radius And welding current waveform of the key factors. The welding experiment was carried out and the training data were collected. The neural network model between welding current, rotation frequency, gun height, torch rotation radius, weld formation and welding current waveform was derived. The final verification test shows that the proposed neural network model can predict the shape of welding seam and the current waveform of welding current well, and can provide the basis for choosing the optimal welding process parameters of rotating arc method to realize the excellent welding process automation.