基于BP网络的冷弯型钢轧后回弹预报模型

Re-spring Forecast Model of Cold Roll-formed Sections Based on BP Neural Networks

运用冷弯型钢的现场实测数据,建立了基于BP网络的轧件轧后回弹预报模型;利用该模型的计算结果,得到回弹角与轧件厚度及轧辊锥角间的表格化模型及显式函数关系式。该模型及函数关系式可用于冷弯型钢的孔型设计。

The re-spring forecast model of workpiece based on BP neural network has been established by using the measured values of cold roll-formed sections. The table model and explicit function expression of re-spring angle with the primitive height and the primitive angle of the roll have also been obtained on the basis of the calculation results of the model. The table model and formula can be used in the pass design of cold roll- formed sections.