BP neural network-based prediction method for rolling force of different-thickness plate

Abstract

The embodiment of the invention discloses a method for predicting the rolling force of a different-thickness plate based on a BP neural network. The prediction method comprises the steps of adopting afinite element software for simulation; rolling force data under different technological parameter conditions are obtained; dividing the obtained rolling force data into two groups; taking the two groups respectively as a training set and a test set of a BP neural network; and determining an optimal BP neural network model by a gradual growth method. According to the BP neural network model, therolling force of the different-thickness plate is predicted, and the prediction error of the rolling force is within 10%.

Description

technical field [0001] The invention relates to the technical field of metallurgy, in particular to a method for predicting the rolling force of differential thick plates based on BP neural network. Background technique [0002] The poor thickness plate manufacturing technology is not only a symbol of the national steel technology level, but also a reflection of the country's industrialization level. In the manufacturing process of differential thick plate, the rolling force is the basis for checking the strength of the rolling mill and optimizing the process. During the rolling of poor thick plates, the speed of the rolls in the vertical direction changes. Therefore, parameters such as the contact arc length and rolling force distribution in the rolling deformation zone are dynamically changed. Traditional rolling theory and geometric relations no longer apply. When rolling a thick plate, the rolling force needs to be calculated accurately, so as to provide a scientific b...

AI Technical Summary

Problems solved by technology

The above three methods for calculating rolling force all have the problems of too many assumptions and large error fluctuations, which cannot meet the industrial demand that the error of rolling force prediction should be within 10%.

Images