Research on Precision Compensation of Cross-Cutting Flying Shear based on BP Neural Network

摘要

In allusion to the numerous nonlinearity factors affecting precision in the cut process of cross-cutting line modeling shear and the difficulties to build precision mathematical model controlled by shear, the essay adopts BP neural network to build model of shear precision compensation system. By Matlab simulation experiment, the data demonstrates nonlinearity complex system can be identified fast and effectively by neural network. The predicted time of the model is 97 ms with the control precision within 0-2 mm hitting 89%, proving that the model can fully fulfil production demands regarding precision and pace and prove that probability that BP neural network is used in precision compensation of cross-cutting line modeling shear.