With the development of robotic welding automation, there is a strong interest in welding seam identification and localization methods with high accuracy, real-time performance, and robustness. This paper proposed a 3D workpiece weld identification and localization method based on DBSCAN (density-based spatial clustering of applications with noise) to realize stable feature extraction for multiple joint types. Firstly, this method employs combinatorial filtering to effectively eliminate non-target point clouds, including outliers and installation platform point clouds, which can minimize the computational load. Secondly, DBSCAN is used to classify workpiece point clouds into different clusters, which can be used for point cloud segmentation of flat workpieces and curved workpieces. Thirdly, the edge detection and feature extraction methods are used to obtain joint gap and weld feature points while combining the information of point clouds for different types of welds. Finally, based on the identification and localization of the welds, welding path planning and attitude planning are implemented. Experimentation results indicated that the proposed method exhibits robustness across various types of welded joints, including butt joints with straight seams, butt joints with curved seams, butt joints with curved workpieces, and lap joints. Meanwhile, the average error of joint gap detection was 0.11 mm and the processing time of a 90 mm straight-seam butt joint is 701.12 ms.
The uncertainties of parameters are inevitable in practical engineering problems, which may cause unexpected variations of objectives and constraints. Tolerance design methods can specify the allowable variation of uncertain parameters, thereby maintaining product quality while achieving lower manufacturing cost. In this paper, a novel tolerance design method is proposed, in which the ‘reverse model’ is used to directly analyse the robustness of the preexisting alternatives and the Kriging model is introduced to reduce computational cost. Firstly, the sensitivity region is generated by the ‘reverse model’ that maps the given acceptable objective and feasibility variations into the parameter space to analyse the robustness. Secondly, interval uncertainty is introduced to describe tolerance within the sensitive region. Thirdly, a nested double-loop structure is adopted to implement the search process. Two numerical examples and one engineering case are used to illustrate the effectiveness of the proposed method. Verification of robustness for the obtained interval is performed by Monte Carlo Method, and the worst-case analysis is applied to verify whether the tolerance is the maximum interval. The obtained maximum tolerance that meets the performance requirements reflects the trade-off between quality and manufacturing cost.
1 April 2021 in Journal of Physics: Conference Series
Porosity is one of the most serious defects in Al alloy laser welding. The online detection of the porosity can identify the weak position of weld seam and take remedial measures accordingly. In this paper, a convolutional neural network (CNN) model where the input is the signal spectrum graphs extracted by wavelet packet decomposition (WPD) is constructed to identify the porosity during Al alloy laser welding in real-time. The porosity monitoring platform is set up to obtain the keyhole opening area signal and the keyhole depth signal in the welding process. The sliding window scanning algorithm is used to scan the signals and the weld seam, and the time-frequency spectrum graphs are obtained by WPD processing on the signals in each sliding window. Through analysis, when there is porosity in a small weld seam section of the current sliding window, the signal spectrum graphs are messy at this moment, while when there is no porosity, the signal spectrum graphs are clean and the frequency bands are concentrated in low-frequency part. The CNN model is constructed to classify the signal spectrum graphs under different porosity status, thereby it can identify the porosity in the weld seam.
22 February 2023 in Structural and Multidisciplinary Optimization