Temperature distribution prediction in control cooling process with recurrent neural network for variable-velocity hot rolling strips

Control cooling is an essential method for microstructure and mechanical property control in hot rolling strip making. Therefore, it vital to realize high-precision temperature distribution prediction process ensure the industrial production. In this paper, a traditional mechanism model based on finite-difference combined with online cycle velocity calculation strategy was introduced as one of baseline methods estimating distribution. However, considering time, variable-velocity makes difficult rapidly modifying water all segments zone. Herein, recurrent neural network proposed, by fully dynamic characteristics. And performance different cell time steps evaluated. The results indicated that proposed could prediction, bi-LSTM 48 timesteps has highest determination coefficient value 0.976, lowest root mean square error 8.03, absolute 5.7. Furthermore, compared model, retained lower computational cost, making applicable application providing real-time prediction.