Real-time recognition of weld defects based on visible spectral image and machine learning

The quality of Tungsten Inert Gas welding is dependent on human supervision, which can’t suitable for automation. This study designed a model for assessing the tungsten inert gas welding quality with the potential of application in real-time. The model used the K-Nearest Neighborhood (KNN) algorithm, paired with images in the visible spectrum formed by high dynamic range camera. Firstly, projecting the image of weld defects in the training set into a two-dimensional space using multidimensional scaling (MDS), so similar weld defects was aggregated into blocks and distributed in hash, and among different weld defects has overlap. Secondly, establishing models including the KNN, CNN, SVM, CART and NB classification, to classify and recognize the weld defect images. The results show that the KNN model is the best, which has the recognition accuracy of 98%, and the average time of recognizing a single image of 33ms, and suitable for common hardware devices. It can be applied to the image recognition system of automatic welding robot to improve the intelligent level of welding robot.

Low-Cost Drive System for Electric Motors based on the IRAM Module

With the increasing use of equipment that demand electric drive systems, the need for new systems that meet requirements of compactness, versatility, safety and low cost has increased. The IRAM module is an electronic circuit that provides a driver for DC and AC motors, being extremely compact and presents high performance. In this context, this work contributes to the power electronics area, presenting a design and construction of a low cost drive system, based on IRAM module, developed for individual or simultaneous drive, up to two DC motors. To carry out the experiments, DC motors responsible for moving a welding robot, were used. Experimental results are presented to shown the feasibility of using this system.

Model Reference Control Applied on the Velocity Tracking of a Welding Robot

Robotics has been expanding over last decades, employed mainly to the activities that are most harmful to human beings. Considering that welding is one of the most risky activities in industries, studies and researches in the process automation are quite important. In this context, this work contributes to the control of the velocity tracking of the displacement of a linear welding robot. The mathematical modelling of the robot is presented, and the chosen control technique is Model Reference Control, which allows project controller based on the desired behaviour for the robot. To corroborate controller effectiveness, simulation and experimental results are presented and discussed, proving that proposed technique is adequate to control the robot velocity.

Dynamic Characteristics Analysis of Two-Beam Laser Welding Robot for Fulselage Panels

Industrial robotics is becoming increasingly popular in the field of manufacturing automation. Two-beam laser welding robot which is a proprietary industrial robot of great importance to improve the welding quality of stringer-skin T-shape structure. In the process of two-beam laser cooperative welding, the robot constantly adjusts its own posture, and the position and posture of each joint would change simultaneously, which leads to the change of the natural frequency, and other dynamic characteristics of the welding robot. Based on the finite element method (FEM), the modal analysis of the robot joints in the range of motion ability and the range of motion in the process of two beam laser welding are studied, which can provide the basis for the design and accurate control of the robot with high degree of freedom (DOF). The dynamic characteristics of the whole robot in different positions and attitudes is carried out, which includes two parts, one is importance ranking of 18 joints of the robot through orthogonal test according to the range of each joint movement. The other is obtaining a plurality of time points in one welding cycle, and performing a modal analysis of the robot at each time point on the basis of the robot joints in the range of motion during the process of two-beam laser welding, the optimal number of time nodes are attained and the test workload could be reduced. The approach described herein provides a theoretical basis for robotics design and control optimization.

Welding seam recognition and tracking for a novel mobile welding robot based on multi-layer sensing strategy

Mobile welding robot with adaptive seam tracking ability can greatly improve the welding efficiency and quality, which has been extensively studied. To further improve the automation in multiple station welding, a novel intelligent mobile welding robot consists of a four-wheeled mobile platform and a collaborative manipulator is developed. Under the support of simultaneous localization and mapping (SLAM) technology, the robot is capable of automatically navigating to different stations to perform welding operation. To automatically detect the welding seam, a composite sensor system including an RGB-D camera and a laser vision sensor is creatively applied. Based on the sensor system, the multi-layer sensing strategy is performed to ensure the welding seam can be detected and tracked with high precision. By applying hybrid filter to the RGB-D camera measurement, the initial welding seam could be effectively extracted. Then a novel welding start point detection method is proposed. Meanwhile, to guarantee the tracking quality, a robust welding seam tracking algorithm based on laser vision sensor is presented to eliminate the tracking discrepancy caused by the platform parking error, through which the tracking trajectory can be corrected in real-time. The experimental results show that the robot can autonomously detect and track the welding seam effectively in different station. Also, the multiple station welding efficiency can be improved and quality can also be guaranteed.

Semi-Supervised Training for Positioning of Welding Seams

Robotic welding often uses vision-based measurement to find the correct placement of the welding seam. Traditional machine vision methods work well in many cases but lack robustness when faced with variations in the manufacturing process or in the imaging conditions. While supervised deep neural networks have been successful in increasing accuracy and robustness in many real-world measurement applications, their success relies on labeled data. In this paper, we employ semi-supervised learning to simultaneously increase accuracy and robustness while avoiding expensive and time-consuming labeling efforts by a domain expert. While semi-supervised learning approaches for various image classification tasks exist, we purpose a novel algorithm for semi-supervised key-point detection for seam placement by a welding robot. We demonstrate that our approach can work robustly with as few as fifteen labeled images. In addition, our method utilizes full image resolution to enhance the accuracy of the key-point detection in seam placement.