1090

In order to realize the prediction of continuous casting billet inclusions and provide guidance for actual production, a slab quality prediction method based on genetic algorithm and random forest (GA-RF) is proposed. Taking the continuous casting production data as the object, the sequential backward selection algorithm (SBS) is used to perform feature selection to remove redundant features, and a random forest algorithm model with genetic algorithm parameter optimization is established to predict the quality of cast slabs. The results show that the accuracy of the GA-RF model in casting billet quality prediction is 89.24%, which is better than the prediction accuracy of 78.25% of SVM and 85.84% of BP neural network. At the same time, compared with the grid search algorithm and the random forest model optimized by the particle swarm algorithm, the GA-RF algorithm has higher prediction accuracy and prediction speed.