Research On Wear State Recognition Of Micro Milling Tool Based On ACO-SVM

In the cutting process,tool wear is an unavoidable phenomenon,and in the process of micro milling technology development,tool wear phenomenon is widely concerned and studied.On the other hand,the development of intelligent automation is constantly bringing forth new ideas,and the development prospect is very good.The development of tool wear condition monitoring technology has also undergone earth shaking changes.In order to ensure the quality and efficiency of modern machine tool production and processing,it is urgent to research and develop more advanced tool wear condition monitoring technology.In this paper,the wear state identification of micro milling tool is studied.The main research contents are as follows:Firstly,the basic situation of micro milling wear and the status quo of tool wear monitoring at home and abroad are summarized.Then,the experimental platform for micro milling is built.According to the experimental equipment used,the vibration signal is taken as the research object,and the experimental system for micro milling tool wear monitoring is established.The vibration signals of micro milling tool in different wear conditions are collected.Then,according to the image of the cutting part of the micro milling tool obtained by the image after the cutter is off-line,the wear state of the micro milling tool in this paper is divided into four kinds.Based on the collected cutting vibration signal data,the appropriate signal processing method is selected,and the time-domain and frequency-domain features are extracted through time-domain and frequency-domain analysis.Then,the feature dimension reduction optimization algorithms of principal component analysis,boundary Fisher analysis and linear discriminant analysis are analyzed.By comparing the visualization results of the low dimension mapping of the three algorithms,linear discriminant analysis is selected as the feature dimension reduction optimization algorithm in this paper.The improved algorithm makes full use of the label information of sample category.Aiming at the problem of high dimension of original feature set,combined with the multi classification of tool state,in order to output the optimal feature reflecting tool wear state,this algorithm is used to reduce the dimension of fused multi feature.Finally,the tool wear state recognition model of micro milling tool is constructed based on support vector machine.The two key parameters of support vector machine are analyzed,and the ant colony algorithm is used to optimize the parameters,and then the ACO-SVM model is established.But the basic ant colony algorithm also has some limitations.After studying the improved algorithm of ant colony algorithm,the recognition model of improved ant colony algorithm optimization support vector machine based on ACS is established.According to the vibration signal collected from the tool wear and the final extracted features,the model is input and evaluated comprehensively.The experimental results also show that the optimized recognition model has comprehensive advantages,and the recognition accuracy has been improved.Therefore,it is feasible to apply the method to tool wear condition monitoring,and has a certain practical value.