Quality prediction of aluminum alloy resistance spot welding based on correlation analysis and SSA-BP neural network

Based on the characteristics of the process signals in the resistance spot welding process, three working conditions of different spacing, different gaps and different spacing and gaps are analyzed, and correlation analysis is introduced to extract the correlation between the process signals and the diameter of nugget. A resistance spot welding quality prediction model based on Sparrow Search Algorithm-Back Propagation Neural Network (SSA-BP) was established, and power, welding current, welding voltage and dynamic resistance are used as input features of the prediction model. The results show that the BP neural network optimized by the sparrow search algorithm outperforms the BP model on the test set with R², MSE, RMSE and MAE of 0.95, 1.55, 1.24 and 0.90, respectively. It is also determined that there exists a mapping relationship between power, welding current, welding voltage and dynamic resistance and the diameter of the nugget, which provides a basis for the design of process parameters for welding.