Real-time K-TIG welding penetration prediction on embedded system using a segmentation-LSTM model

Keyhole tungsten inert gas(K-TIG)welding is capable of realizing single-sided welding and double-sided forming and has been widely used in medium and thick plate welding.In order to improve the accuracy of automatic weld identification and weld penetration prediction of robot in the process of large workpiece welding,a two-stage model is proposed in this paper,which can monitor the K-TIG welding penetration state in real time on the embedded system,called segmentation-LSTM model.The proposed system extracts 9 weld pool geometric features with segmentation network,and then extracts the weld gap using a traditional algorithm.Then these 10-dimensional features are input into the LSTM model to predict the penetration state,including under penetration,partial penetration,good penetration and over penetration.The recognition accuracy of the proposed system can reach 95.2%.In this system,to solve the difficulty of labeling data and lack of segmentation accuracy,an improved LabelMe capable of live-wire annotation tool and a novel loss function were proposed,respectively.The latter was also called focal dice loss,which enabled the network to achieve a performance of 0.933 mloU on the testing set.Finally,an improved slimming strategy compresses the network,making the segmentation network achieve real-time on the embedded system(RK3399pro).

Investigation into keyhole-weld pool dynamic behaviors based on HDR vision sensing of real-time K-TIG welding process through a steel/glass sandwich

To obtain a deep insight into keyhole tungsten inert gas welding,it is necessary to observe the dynamic behavior of the weld pool and keyhole.In this study,based on the steel/glass sandwich and high dynamic range camera,a vision system is developed and the keyhole-weld pool profiles are captured during the real-time welding process.Then,to analyze the dynamic behavior of the weld pool and keyhole,an image processing algorithm is proposed to extract the compression depth of the weld pool and the geometric parameters of the keyhole from the captured images.After considering the variations of these parameters over time,it was found that the front and rear lengths of the keyhole were dynamically adjusted internally and had opposite trends according to the real-time welding status while the length of the keyhole was in a quasi-steady state.The proposed vision-based observation method lays a solid foundation for studying the weld forming process and improving keyhole tungsten inert gas welding.