Prediction and analysis of slab quality based on neural network combined with particle swarm optimization (PSO)

Abstract

Based on the study of the mechanism of bloom crack, the main factors affecting the quality of bloom are determined. The intelligent optimization algorithm combining PSO and Back Propagation(BP) neural network is introduced to establish the prediction model based on typical defects. Collect on-site sample data, normalize it, and PSO is used to recalculate the weights and thresholds to accelerate the convergence and improve the accuracy and stability of the results. The experimental results show that the prediction accuracy of the optimized neural network model is high, and it is closer to the actual production of continuous casting.

Keywords

• continuous casting
• steel slab quality
• neural network
• particle swarm optimization
• prediction model