Prediction for surface roughness of the large-pitch internal thread based on homologous isomerism data

It is difficult to measure the surface roughness of large-pitch internal threads; predictions are used instead measurements, whereas common not highly accurate and narrow in scope use. The homologous isomerism data vibration signal were utilized establish a predictive model, which predicted threads. corresponding acquired by turning thread, processed using Relief-F algorithm obtain weights, effects different features on roughness. Additionally, influenced structural characteristics workpiece with large pitch small number teeth, support vector machine (SVM) radial basis function neural networks (RBF-NNs) model as input parameters. Eventually, SVM higher accuracy prediction better ability generalization was more appropriate for research this paper through comparison analysis. verified that absolute error less than 0.05 μm, relative 4% both left right threaded surfaces, demonstrating could take place measuring mass production method proposed can also be extended other parts roughness, especially those whose due characteristics.