Research on the prediction model for recovery rate of alloying elements

It is well known that alloying model has great effect on steel quality, and the precision of alloying model relies heavily on the calculation of recovery rate of alloying elements. Aiming at steel quality improvement, this paper firstly built a recovery rate prediction model with BP neural network, and then worked on this model with the help of LM and POS algorithms respectively. The comparison of simulation shows that, the PSO algorithm can overcome the shortcomings of local minimum and improve the precision of convergence to a certain extent. The simulation results confirmed the high efficiency of this algorithm.