How to Accurately Monitor the Weld Penetration From Dynamic Weld Pool Serial Images Using CNN-LSTM Deep Learning Model?
How to Accurately Monitor the Weld Penetration From Dynamic Weld Pool Serial Images Using CNN-LSTM Deep Learning Model?
复制标题
DOI：
10.1109/lra.2022.3173659
复制
发表时间：
2022-07
影响因子：
5.2
通讯作者：
Rui Yu;J. Kershaw;Peng Wang;Yuming Zhang
中科院分区：
文献类型：
--
作者：
Rui Yu;J. Kershaw;Peng Wang;Yuming Zhang
这封信说明了如何通过确保原始信息的充分性和使用适当的深度学习网络来提取相关信息来解决一个具有挑战性的问题。所涉及的问题是准确监测完全熔透的熔池中的熔深，该熔深由工件背面的焊缝宽度来量化。这是具有挑战性的，因为穿透发生在工件表面之下，不可见。一种流行的方法是使用熔池图像来得出它。对物理过程的分析表明，单个熔池不包含足够的信息，但最新的连续熔池可能包含足够的信息。因此，尽管深度学习模型可能会提取已经存在的信息，但原始信息可能还不够。因此，需要一种能够从动态连续熔池图像中提取信息的模型。为此，提出了一种CNN-LSTM(卷积神经网络与长-短期记忆相结合)模型。通过随机改变焊接电流和焊接速度，实验产生了动态熔池。在实验过程中，使用HDR摄像机对熔池进行成像。还可以从工件的背面捕获图像，为培训、验证和测试提供基本事实。结果表明，在焊缝背面宽度为0.3 mm的连续熔池图像中，可以准确地预测高度动态变化的熔池。与比较研究的结果进行了比较，以验证信息充分性(通过使用连续图像)和特征提取能力(通过使用深度学习)的有效性和贡献。
This letter illustrates how a challenging problem be solved by assuring the adequacy of the raw information and using an appropriate deep learning network to extract the relevant information. The problem concerned is accurate monitoring of the penetration, in a fully penetrated weld pool, as quantified by the width of the weld on the back-side of the workpiece. This is challenging as the penetration occurs below the workpiece surface and is not visible. A popular method is to use a weld pool image to derive it. Analysis of the physical process suggests that a single weld pool does not contain adequate information but most recent serial weld pools may. As such, although a deep learning model may extract information that is already there, the raw information may not be sufficient. Hence, a model that is capable of extracting information from dynamic serial weld pool images is needed. To this end, a CNN-LSTM (convolutional neural network combined with long-short term memory one) model is proposed. Dynamic weld pools are experimentally generated by changing the welding current and welding speed randomly. The weld pools are imaged using an HDR camera during experiments. Images are also captured from the back-side surface of the workpiece to provide the ground truth for training, validation, and testing. It is found that the highly dynamically changing weld pool can be accurately predicted using serial weld pool images at 0.3 mm for its back-side bead width. Comparison has been made with results from comparative studies to verify the effectiveness of and the contribution from the information adequacy (by using serial images) and the feature extracting capability (by using deep learning).