基于特征加权的朴素贝叶斯主轴热误差建模

Thermal error modeling of spindle based on feature weighted Naive Bayes

采用基于特征加权的朴素贝叶斯方法,对铣削电主轴热误差进行了建模及预测研究。根据不同特征对热变形影响程度的不同,采用信息增益的方法计算了不同特征的权重,基于实验得到的数据,建立了轴径向热变形误差模型,从预测结果看,模型具有比较好的预测功能。

Method of feature weighted Nave Bayes is used for modeling and forecasting thermal error of electrical spindle. According to the influence on thermal deformation,different weights of characteristics are calculated using information gain method. The error model of the thermal deformation is established based on the experimental data. The prediction results show that the model has a good prediction function.