Pressure Prediction Model in Transition Ladle based on Adaptive PSO-BP Neural Network

Generally, the pressure in the transition ladle during the manufacturing of amorphous alloys could be influenced by numerous factors. Given these issues, BP neural network is employed in this paper to ensure the prediction of the pressure in the transition ladle in the production of amorphous alloys, and adaptive PSO is adopted to adjust the weights and thresholds. The particle swarm is adaptively divided by the median value, the best location is searched locally, and the flight direction and speed of each particle are adjusted to obtain good results. Finally, emulations are performed using the amorphous alloy production site data of a steel plant. The results demonstrate that the pressure prediction model has higher prediction precision in the transition ladle established by the adaptive PSO-BP neural network. This can lay a foundation for the latter pressure control settings in the transition ladle.