Research on Spark Splash Control of Resistance Welding Based on Big Data Analysis Algorithm

Welding is a key process in automobile body manufacturing, and its quality has an important impact on the connection strength and paint quality of vehicle body. It is worth noting that long-distance spark splashing during welding will not only damage the body paint, but also pose a threat to the personal safety. It is found that welding parameters are the direct factors affecting spark posture and splash distance. Therefore, the selection and optimization of welding parameters is very important. Aiming at the control of spark splashing distance during welding, this paper selects the welding process parameters that have the greatest impact on spark posture and splash distance: welding current and welding time, and uses fully connected neural network, Bayesian regression and Gaussian mixture model to mine the implicit relationship between spark splashing distance and the selected welding process parameters, the optimal welding process parameters can be deduced through the hidden relationship, so as to improve the quality, efficiency and safety factor of body welding.