Research on MLD Modeling Method of Robot Aluminum Alloy Pulsed TIG Welding Process Based on Visual Sensing

This study focuses on the hybrid logic dynamic (MLD) modeling method of robot aluminum alloy pulse TIG (Tungsten Inert Gas) welding process based on vision sensing. Welding is a key technology in modern manufacturing industry, especially in the application of aluminum alloy materials, and its welding quality control is facing complex challenges. This study aims to improve the automation and intelligence level of welding process by combining robot technology and visual sensing technology. A data acquisition system based on high-precision vision sensor and advanced image processing algorithm is designed, which realizes real-time and accurate acquisition of weld and molten pool information. The weld contour is identified by machine learning algorithm, and the geometric parameters and dynamic change characteristics of weld pool are extracted, and the statistical relationship model between welding parameters and geometric characteristics of weld pool is established. On this basis, the MLD model is constructed, which can describe the continuous and discrete dynamics in the welding process. The experimental results show that MLD model can accurately predict the change trend of weld pool geometric parameters, effectively identify the welding state, and realize accurate control of welding process. This study not only improves the stability of welding quality, but also promotes the further development of aluminum alloy welding technology, which has important practical application value and broad application prospects.