Image and Vibration Based Mixed Variable Approach For Machining Performance Estimation

摘要

Automation from design to final product poses several challenges in terms of addressing geometry, dimensional and tolerances requirements of products. While geometry and dimensional requirements can be addressed through one time interfacing between CAD, CAM and machining systems, addressing surface quality requirement needs real time monitoring and control. In this work, images along with the vibration signature from accelerometer measured during machining are used for monitoring the metal cutting process. Data from the image processing unit and the accelerometer are incorporated into a mixed variable estimation approach for evaluation of machining performance like tool wear and surface roughness. Support Vector Machine (SVM) model is developed for estimation of machining performance. The results are encouraging and this method can be implemented in real time estimation systems.