Visual Neural Network Model for Welding Deviation Prediction Based on Weld Pool Centroid

Abstract

To solve the problem in the process of weld seam tracking, a new prediction model for welding deviation based on the weld pool image centroid has been proposed in the paper. First, some weld images under different weld currents were captured by a vision sensor. A composite filter system, which is composed of narrow-band and neutral filters, is used to reduce the disturbance of weld arc. So, several clear weld pool images can be obtained. Then a frontier of weld pool is chosen to be the processing region. Median filter and gray transformation operations are used to enhance the contrast of processing region. On this basis, the variation trend of centroid difference [Formula: see text] and welding deviation [Formula: see text] were analyzed. The centroid difference value [Formula: see text] and the weld current [Formula: see text] were determined to be welding status parameters. Moreover, a BP neural network was set up, which was composed of three layers. Next, elastic gradient descent method was used to be the training function. So a prediction model between the welding status parameters [Formula: see text] and [Formula: see text] and the welding deviation [Formula: see text] was set up. In the end of the paper, several experiments were performed to test the accuracy of the setup prediction model. The results showed that prediction values of welding deviation calculated by the vision model are fit to the real measured values. The final errors of the vision model under the weld current 70[Formula: see text]A and 73[Formula: see text]A were 0.033[Formula: see text]mm and 0.027[Formula: see text]mm, which showed excellent accuracy, environmental suitability and intelligence of the model.

Compressive Strength Prediction of High-Performance Hydraulic Concrete Using a Novel Neural Network Based on the Memristor

“…The learning algorithm is an extremely important part of BP neural network. In order to improve the performance of neural network, many other advanced algorithms have been consecutively proposed, such as the gradient descent with momentum (GDM) [32], the gradient descent with adaptive learning rate (GDA) [33], the elastic gradient descent (EGD) [34], and the Levenberg-Marquardt algorithm (L-M) [35]. These five classical algorithms are also utilized to compare with the proposed model in this study.…”

Robotic Weld Seam Correction Control System Based on Visual Sensing

“…发生变化，以致焊缝与机器人示教路径有偏差 [3][4] 。 对于该问题较好的解决方法是能够根据焊缝的变形 信息，实时观察并能够做出调整以纠正焊枪路径， 从而保证焊接质量 [5] 。实现机器人焊接路径的实时 纠偏，一般需要三个过程：实时焊缝信息的获取、 对焊缝信息的分析处理并发出相应的控制信号、机 器人接收信号并做出调整。通常，根据反馈信息类 型的差别，机器人视觉伺服一般分为基于位置的视 觉伺服控制、基于图像的视觉伺服控制和混合视觉 伺服控制 3 类 [6][7][8] 。LI 等 [9] 采用的是利用结构光视觉 来设计自动焊缝跟踪和识别。DING 等 [10][11] 通过观 察焊缝中心的焊接预测偏差建立了视觉神经网络模 型，有助于焊缝跟踪的发展。PRECUP 等 [12]…”

Weld seam tracking for TIG welding of stainless steel pipes based on machine vision

“…In order to improve the process quality of welding products and improve the efficiency of welding production, domestic and foreign scholars have conducted a lot of research on weld tracking technology and welding process. Ding et al proposed a welding deviation prediction model based on the centroid of weld pool image to solve the problems in the welding seam tracking process [3] . By analyzing the changing trend among centroid deviation of weld pool, welding current and welding deviation, and combining with BP neural network technology, the prediction model of welding deviation was established.…”