Research On Nc Tool-wear Detection Based On Machine Vision

In the milling process,regularly grasping the changes in the machining state of the tool and replacing the failed tool in time can reduce the downtime caused by the tool failure and significantly improve the machining quality.Compared with traditional manual inspection and offline instrument inspection,the on-machine inspection method of end mill wear based on machine vision effectively avoids installation errors and inspection delays caused by loading and unloading tools,which is conducive to improving inspection efficiency and inspection accuracy;the current on-machine vision Most of the detection methods build the detection device inside the CNC machining center,which realized on-machine detection to a certain extent,but doesn’t fully consider the problems of processing interference and lens contamination.In order to solve the above problems,based on the linear transmission mode of the telescopic diamond linkage mechanism,this paper proposes an on-machine vision detection scheme for the wear of end mills built outside the side observation window of the machining center.The image acquisition device is driven by the link mechanism to enter or exit the CNC machine tool.Through the steps of 3D model design,parts selection,processing and assembly,the detection device is built,and the control system scheme is designed in combination with the on-machine inspection process of end mill wear,so as to realize on-machine inspection of end mill wear that is relatively independent from the outside of the machining center,effectively solving the problems of machining interference and lens contamination.For the research on the wear detection algorithm of end mills,the detection system uses adaptive hybrid filter denoising,contrast enhancement and other preprocessing,and uses the mean iteration method and the maximum inter-class variance method to perform image segmentation to obtain a binary image of the worn area.The Canny operator and the sub-pixel edge based on the Zernike moment are fused to extract the wear contour,the end mill wear contour model is created based on the image plane,and the original cutting edge is reconstructed with the least square method.The wear amount of the end mill is obtained and the wear state of the end mill is fed back,which realized the rapid detection of the end mill wear.The control software of the on-machine inspection system integrates the functions of mechanism motion control,end mill image acquisition and processing,and CNC communication control.The movement of the detection mechanism is controlled by PLC;The end mill wear detection is completed through the image acquisition and processing analysis of the end mill;The end mill wear compensation feedback is completed by communicating with the CNC system.Finally,five end mills were selected to carry out milling experiments,and the measurement results of the detection system and Tianzhun-VMC322 image measuring instrument were compared.The results show that the measurement deviation of the detection system is less than0.01 mm,and the comprehensive average accuracy rate reaches 96%.