Monitoring Of Laser Welding Process Research Articles

Deep-learning-based porosity monitoring of laser welding process

A porosity monitoring scheme for laser welding process was developed based on a deep learning approach. The in-process weld-pool data were sensed with a coaxial high-speed camera and labelled with the porosity attributes measured from welded specimens. A convolutional neural network (CNN) model with compact architecture was designed to learn weld-pool patterns to predict porosity. In laser welding experiments of 6061 Aluminum alloy, the CNN-based monitoring model achieved a classification accuracy of 96.1% for porosity occurrence detection, though the prediction of micro (less than 100 µm) and deep subsurface pores still remains challenging.

A Multi-Sensor Data Fusion System for Laser Welding Process Monitoring

Most existing laser welding process monitoring (LWPM) technologies focus on detecting post-process defects. However, in sheet metal laser welding applications such as welding of electronic consumer products during mass production, in-process defect detection is more important. In this article, a compact LWPM system using multi-sensor data fusion to detect in-process defects has been built. This system can collect the time series of plasma intensity, light intensity and temperature data for feature analysis. To verify the system's effectiveness, a plasma-light-temperature dataset has been compiled, which consists of 5,836 samples of nine classes, including one positive class and eight negative classes of typical in-process defects. A multi-sensor data fusion network based on a convolution neural network for in-process defect detection, called IDDNet, has also been proposed. Experimental results have demonstrated that IDDNet can achieve better multi-classification results than the support vector machine, with an overall accuracy of 97.57%. In particular, considering this monitoring process as a binary classification problem, IDDNet can achieve a 99.42% accuracy. Moreover, IDDNet can reach an average speed of 0.79ms per sample on a single GTX 1080ti graphics card, which meets the real-time requirement for industrial production. The proposed LWPM system has been successfully verified in real applications of sheet metal laser welding.

WPD-PCA-Based Laser Welding Process Monitoring and Defects Diagnosis by Using FNN and SVM

Recent years have seen increasing attention paid to laser welding monitoring. This paper introduces an innovative approach to perform laser welding process monitoring and welded defect diagnosis. The laboratory-scale sensor can be replaced with industrial-scale sensors after the data-driven model has been established by applying multivariate statistics and machine learning methods. In addition, industrial-scale sensor makes effective diagnosis of welded defect by using pattern recognition. Experimental results show that the feature vector affecting estimation and classification accuracy can be obtained by using wavelet packet decomposition principal component analysis. Image processing technology was applied to quantify geometrical parameters of welding process. The feedforward neural network prediction model and the support vector machine classification model built in this research help to guarantee accurate estimation on welding status and effective identification of welded defect. The method proposed by this paper provides an innovative data-driven-based approach for laser welding process monitoring and defects diagnosis.

Non-destructive Real Time Monitoring of the Laser Welding Process

Laser welding is a high power density technology of materials joining that has many advantages in comparison with conventional fusion welding methods, for example, high accuracy, flexibility, repeatability and especially very narrow heat-affected zone which results in minimal workpiece distortions. Since it is still quite expensive technology, minimal spoilage is required. Effective system of quality control and processing parameters optimization must be established to reduce total costs, which is particularly required in industrial production. In this article some results of pulsed Nd:YAG laser welding process monitoring based on the measurement of plasma electron temperature are presented. The ability of designed sensor to detect weld penetration depth has been demonstrated. Plasma spectral lines intensities measurement can discover gap instabilities as well as local sheet thickness reduction.

A Novel PSO Algorithm Model Based on Population Migration Strategy and its Application

According to the intelligent behavior of social population, the centroid of the individual best of particle swarm is firstly introduced in p article s warm o ptimization (PSO) model to enhance inter-particle cooperation and information sharing capabilities, then combining the m echanism of p opulation m igration a lgorithm (PMA) , a novel PSO algorithm with adaptive space mutation (PM-CPSO) is proposed to im prove computing performance of PSO algorithm. Experiment results of Benchmark function and practical application in quality monitoring of laser welding process show the new algorithm has not only higher convergence precision and faster convergence speed, but also can avoid the prema ture convergence problem effectively.

Nd:YAG laser welding process monitoring by non-intrusive optical detection in the fibre optic delivery system

This paper describes a non-intrusive optical sensing technique for Nd:YAG laser welding monitoring. The cladding power monitor (CPM) detects the light returned through the cladding of the delivery fibre optic, and appropriate spectral filtering isolates the light radiated from the plume of hot gas which forms above the weld. The successful detection of welding faults using this optical signal is described, including laser focus errors and shield gas interruption.