Welding process expert system based on Industrial Internet and neural network

Welding process is a nonlinear, multi-parameter coupling and complex process that is difficult to quantify accurately. This paper innovatively combines Industrial Internet technology and neural network technology to establish a welding process expert system. Through the use of Industrial Internet, the welding process expert system automatically collects the process data generated by the welding equipment in the process of operation, and uploads these process data to the cloud platform. In the cloud platform, the system uses welding process parameters to conduct neural network learning, establish prediction models, realize high-precision prediction of mechanical properties of welded joints, and achieve the purpose of reducing the number of welding process qualification tests.