基于支持向量机的钛合金表面粗糙度预测分析

Prediction analysis of titanium alloy surface roughness based on support vector machine

为了准确预测钛合金丝材的表面粗糙度,设计了无心车床切削TC4钛合金丝材试验,对不同切削参数下无心车床的前导向机构、主轴机构、后导向机构的8通道振动参数进行测量,采用皮尔逊相关系数对测量点振动参数进行特征选择和特征降维,建立支持向量机(SVM)的表面粗糙度预测模型。在不同的主轴转速、进给速度等工艺参数下,得到SVM预测模型的表面粗糙度的预测精度:RMSE为0.0268,MAPE为0.0403,R^(2)为0.8274,基于SVM模型预测钛合金线材的表面粗糙度具有较好的精度,验证了模型的有效性。

In order to accurately predict the surface roughness of titanium alloy wire,experiments on cutting TC4 titanium alloy wire by centerless lathe are designed.The 8-channel vibration parameters of the front guide,spindle mechanism and back guide mechanism of the centerless lathe are measured under different cutting parameters.The surface roughness prediction model of support vector machine(SVM)is established by using Pearson correlation coefficient for feature selection and dimension reduction of vibration parameters of measuring points.The prediction accuracy of support vector machine model is verified by measuring experiment under different spindle speed and feed speed.The prediction accuracy of SVM prediction model is RMSE=0.0268,MAPE=0.0403,R^(2)=0.8274.The results show that the support vector machine model has a good precision in predicting the surface roughness of titanium alloy wire,which proves the validity of the model.