An integrated prediction model of sinter comprehensive performance

The process of sinter have a great influence on production and economic benefits of the iron-making, the prediction of sinter performance is very necessary. However, the sinter process is complex, nonlinear and hysteresis, and strong coupling, it´s impossible to establish a mathematical model. In this paper, we constructed a prediction model of sinter comprehensive performance by integrating GM(1,1) and BPNN based on information entropy since a single model can´t suit different sinter quality. The application results show that the prediction model has high accuracy rate, stability and efficiency.