基于深度学习的白车身焊接路径智能规划方法

Welding Path Planning of Body-in-White Based on Deep Learning

针对白车身焊接路径规划经验规则多、路径复杂等问题,提出一种端到端的序列到序列深度学习框架,通过学习大量历史数据来提高路径规划的效率与质量。基于LSTM编码器、LSTM解码器和注意力机制搭建深度学习模型,对焊点数据进行归一化和预排序,同时构建了焊点表,提升了模型的训练效率和预测准确率。实现了向模型输入无序的焊点序列,输出目标焊点序列。在白车身左前门焊接生产线上的实验结果表明,所提方法相较于经典机器学习算法,在面对多样的约束条件时准确性更高,焊接路径规划结果更合理。

Aiming at the problem of too many empirical rules and complex paths in the welding path planning of the body-in-white,thispaperproposes an end-to-end sequence-to-sequence deep learning framework to improve the efficiency and quality of path planning by learning a large amount of historical data.A deep learning model is built based on LSTM encoder,LSTM decoder and attention mechanism.The weldpoint data is normalized and pre-sorted,and a weld point table is built to improve the training efficiency and prediction accuracy of the model.The target weld point sequence is obtained from unordered weld point sequence.The experimental results on the welding production line of the left front door of the body-in-white show that the proposed method is more accurate when facing various constraints,and the welding path planning results are more reasonable.