Determining surface roughness of machining process types using a hybrid algorithm based on time series analysis and wavelet transform

In this paper, the surface roughness of the machined workpieces are estimated using a hybrid algorithm based on time series analysis and wavelet transform and differences between contact method with this method are presented. The suggested method is based on the exact recognition of the surface dynamic properties in lapping, external grinding, flat grinding, turning, and horizontal milling processes using the largest Lyapunov exponent parameter in time series analysis. This method has the unique ability to reduce image noise and remove image curvature which are produced due to reflection of light or surface geometric curvature in the mentioned machining process. Also, in order to select the appropriate length of the captured images and increasing accuracy of estimating the surface roughness, the image entropy criterion in image processing is used. The results show that the estimated surface roughness are considerably close to the measured surface roughness by the contact method.