基于BP神经网络预测筒形件对轮旋压内径扩径量

2018-06-20

作者：郭代峰，赵俊生，贺胜，史永鹏，吉梦雯单位：中北大学

摘要：内径扩径量是衡量筒形件对轮旋压好坏的关键性指标。本文采用单因素试验设计方法和数值模拟技术，获得了以减薄率、进给比、圆角半径为试验因素，内径扩径量为评价指标的模拟数据。基于试验因素和评价指标，运用BP神经网络技术建立了3-10-1的三层神经网络结构预测模型。用模拟试验所得到的数据对该模型进行了训练和预测，将预测值与实测值相比较，证明该模型对筒形件的内径扩径量有很好的预测效果。

关键词：内径扩径量；单因素试验法；筒形件对轮旋压；BP神经网络

中图分类号：TG376；TH162文献标志码：A

Prediction on Wheel Spinning Diameter Expanding of Volume Shaped Cylinder Based on BP Neural Network

Guo Daifeng，Zhao Junsheng，He Sheng，Shi Yongpeng，Ji Mengwen

Abstract:The diameter expanding volume of cylindrical parts is a measure of the key indicators for the spinning quality.This paper uses the single factor experimental design method and numerical simulation technology are obtained using the thinning ratio，feed ratio，radius as experiment factors，diameter expanding volume data for the simulation evaluation index.Based on the test and evaluation factors index，set up three layer neural network prediction model of 3-10-1 using BP neural network technology.The training and prediction of the model are obtained by simulation test data，the prediction is compared with the measured values，the results show that the model of expanding diameter cylindrical piece size has a good prediction result.

Keywords:inner diameter expansion；single factor test；barrel spinning；BP neural network