A neural network prediction analysis of breakout continuous casting based on differential evolution (DE)

In order to predict the breakout of continuous casting accurately and timely, a breakout prediction neural network (BPNN) model based on differential evolution algorithm was proposed. Differential evolutionary algorithm is introduced to solve the problem of fast optimization. The pretreatment of weights and thresholds improves the accuracy of breakout. The experimental analysis shows that the convergence speed of DE-BPNN breakout prediction model is faster than that of traditional neural network, and the recognition ability are significantly improved