Learning
Welding is a process for connecting parts to each other, which is of great importance, and the presence of any defect and discontinuity in the weld reduces the quality of the weld and lowers the efficiency of the connection. In this article, we examine the important parameters for proper welding and classify the weld texture into two groups of high quality and poor quality using machine learning algorithms(Figure1).
The Importance of Welding Texture Quality Control
Welding quality assurance is critical in ensuring the dependability, durability, and safety of welded structures and components in industrial applications. Adhering to quality standards and implementing effective inspection processes is critical for preventing defects, maintaining structural integrity, and meeting regulatory requirements.
What Is a Weld Defect?
A weld defect results from a poor weld, weakening the joint. It is defined as the point beyond the acceptable tolerance in the welding process. Imperfections may arise dimensionally, wherein the result is not up to standard. They may also take place in the form of discontinuity or in material properties. Common causes of welding defects come from incorrect welding patterns, material selection, skill, or machine settings, including welding speed, current, and voltage.
When a welded metal has a welding defect present, there are multiple options for resolving the issue. In some cases, the metal can be repaired, but at other times the metal itself has melted and the welding procedure needs to be restarted.
Weld irregularities occur for a variety of reasons and it results in different welding defects. They can be classified into two major categories: internal welding defects and external welding defects.
Slag and other entrapments, cracks, microfissures, porosities, inclusions, lack of penetration and fusion, oxide and other scales, are exemplification of welding defects, which deteriorate the corrosion properties of weld metal. These defects normally cause early pitting and crevice attacks in the weld metal.
Welding Quality Control System
First, using the vision system shown in Figure 2, the texture of regular and irregular welds are imaged Then. To adapt the system, machine learning algorithms are used to identify images and categorize them into two groups of low quality and high quality.
What is Machine Learning?
Machine learning is a branch of artificial intelligence (AI) and computer science which focuses on the use of data and algorithms to imitate the way that humans learn, gradually improving its accuracy. Machine learning is an important component of the growing field of data science. Through the use of statistical methods, algorithms are trained to make classifications or predictions, and to uncover key insights in data mining projects.
How Machine Learning Works?
1.A Decision Process: In general, machine learning algorithms are used to make a prediction or classification. Based on some input data, which can be labeled or unlabeled, your algorithm will produce an estimate about a pattern in the data.
2.An Error Function: An error function evaluates the prediction of the model. If there are known examples, an error function can make a comparison to assess the accuracy of the model.
3.A Model Optimization Process: If the model can fit better to the data points in the training set, then weights are adjusted to reduce the discrepancy between the known example and the model estimate. The algorithm will repeat this “evaluate and optimize” process, updating weights autonomously until a threshold of accuracy has been met.
Vision System with using Machine Learning
In this article, by using machine learning, the system is adapted in such a way that by receiving an image of the weld texture, it is possible to report the high quality or poor quality of the weld.(Figure3)
Conclusion
A method for texture identification was proposed to help the visual inspection of weld beads using a passive camera already present in an automated linear welding system. It was based on the premise that an irregular surface texture could be associated with welding internal defects. The classifier used in the developed algorithm showed an accuracy of 96.4% accuracy.