Research On Monitoring The Wear State Of Milling Tool Based On Acoustic Emission And Machine Vision

In machining process,the remaining life of the tool is usually judged by the subjective experience of the worker and the tool is replaced.Using this traditional method will often be difficult to grasp the tool change time point,thus increasing the machine downtime and reducing tool utilization.The key to improve the tool utilization rate and production efficiency is to accurately grasp the tool wear state,timely replace the tool that has reached the life and ensure the full use of the tool.Therefore,Tool Condition Monitoring(TCM)technology was developed to automatically monitor tool wear status.reminding the staff to change the knife could effectively reduce the downtime in processing,thus controlling the time and cost of processing.This paper uses machine vision and Acoustic Emission(AE)signals to monitor tool wear status.On the one hand,the machine vision method in the direct method is used to process the tool wear image to directly extract the wear state and wear amount of the tool.On the other hand,the tool wear state model is established by extracting the characteristics of the acoustic emission signal generated during the machining process,thus judging the wear status of the tool indirectly.The main research contents of this paper are as follows:(1)According to the tool wear form,wear process and international general blunt standards,the average wear width VBave and the maximum wear width VBmax of the flank face are purposed as the blunt standards for nickel-based alloys in this paper.The tool flank wear area AVB is verified that it could reflect the degree of tool wear.A tool wear monitoring platform is designed to realize the collection of tool wear images and acoustic emission signals during processing;(2)The machine vision monitoring system is designed and the tool images are collected through a charge-coupled device(CCD)camera.Then,the tool image is selected through the Structural Similarity Index(SSIM)to realize the automatic image acquisition,Harris corner detection is used to realize automatic cropping of Region of Interest(ROI).Finally,a series of image processing methods are used to extract geometric parameters of tool wear such as VBave,VBmax and AVB from the tool image;(3)The milling tests are carried out with nickel-based superalloy GH4169 as the workpiece.The wear amount calculated by the visual monitoring system is compared with the real wear amount to verify the accuracy of the system.Then,a full-life milling test is carried out,acoustic emission signals that change with the degree of tool wear during the machining process are collected.Finally,the tool wear status is judged through the visual monitoring system;(4)The acoustic emission signals are analyzed in time domain and frequency domain to extract features related to tool wear.The tool wear status model is established through Genetic Algorithm(GA),Particle Swarm Optimization(PSO)algorithm and Support Vector Machine(SVM)to effectively identify the tool wear status.