Product quality prediction of rolling mill in big data environment

With the wide use of rolling mill in iron and steel industry, the quality of rolling mill products has become the primary goal of people. However, due to design defects and manufacturing quality problems, the quality of steel products is seriously affected, and the surface roughness and thickness of steel plate are important quality indicators. In this paper, by analyzing a large number of monitoring data of rolling mill condition and using BP neural network model [1], the discrete system model between monitoring data and “surface roughness” and “thickness error” of rolling steel plate is further established.