采用BP神经网络预测碳纤维增强树脂基复合材料的钻削力

发布时间：2018-04-15 13:33

本文选题：碳纤维复合材料 + BP神经网络　；参考：《机械科学与技术》2017年04期

【摘要】：采用双锋角钻头对碳纤维复合材料进行钻削试验,基于反向传播算法的人工神经网络建立钻削轴向力与主轴转速、进给速度之间的非线性关系模型,对比分析三种不同第二主切削刃与第一主切削刃之比的双锋角钻头在试验加工参数下钻削轴向力变化规律。结果表明:与多元线性回归预测模型对比,在相同试验数据为基础的预测计算下,BP神经网络预测值相对误差明显减小,网络预测值误差均在3%之内,而多元线性回归模型最大误差值达到了12.46%,BP神经网络能建立更精准轴向力预测模型。从降低钻削轴向力的角度分析,应采用第二主切削刃与第一主切削刃之比为1的双锋角钻头进行钻削加工。
[Abstract]:The experiment of drilling carbon fiber composite material with double angle drill bit is carried out. Based on the back propagation algorithm, an artificial neural network is used to establish a nonlinear model of the relationship between the axial force of drilling and the speed of spindle and feed speed.This paper compares and analyzes the variation of axial force of the double-front bit with three different ratio of the second main cutting edge to the first main cutting edge under the experimental processing parameters.The results show that, compared with the multivariate linear regression prediction model, the relative error of BP neural network is obviously reduced under the prediction calculation based on the same experimental data, and the error of the network prediction value is within 3%.The maximum error of the multivariate linear regression model reaches 12.46% BP neural network can establish a more accurate prediction model of axial force.From the point of view of reducing the axial force of drilling, a double-front angle bit with the ratio of the second main cutting edge to the first main cutting edge should be used for drilling.
【作者单位】：湖南科技大学机电工程学院;机械设备健康维护省重点实验室;
【基金】：国家自然科学基金项目(51275168) 国家科技重大专项项目(2012ZX04003031)资助
【分类号】：TP183;TB332

【相似文献】