Tool Wear State Recognition Of Lathe Based On Accelerated Subspace Iteration Method And E Xtreme Gradient Boosting Tree

CNC machine tools are the most important basic equipment in the manufacturing industry,and metal cutting is also one of the important processing methods in the manufacture industry.Usually in the machining process of CNC machine tools,tool wear is inevitable.Therefore,it is necessary to monitor the wear condition of the tool during the machining process.The research and analysis of the tool wear state has important significance for monitoring the machining state of the tool and improving the production quality.This paper mainly focuses on the research of tool wear state recognition during CNC lathe machining.In this paper,the accelerated subspace iteration method and the eXtreme Gradient Boosting Tree algorithm are applied to the research of tool wear state recognition.This method can recognition the tool wear state more accurately and more quickly.In this paper,the vibration signal generated during the cutting process of the tool is selected as the research object.The tool wear state recognition experiment and signal acquisition system are built,and the vibration signal is collected.Firstly,the wavelet packet decomposition and reconstruction method is used to denoise the directly collected signal,then the signal is framed,and the processed signal segment is analyzed by time domain and frequency domain respectively.The data set of multiple statistical variables in the wear state is used to reduce the dimensionality of the statistical variable data set by the accelerated subspace iteration method,and extract the feature information with strong correlation with the tool wear state.Then the processed data set is divided into training data sets and test data sets of different wear stages according to a certain proportion.The eXtreme Gradient Boosting Tree algorithm is used to classify and identify the wear state of the tools,output the classification results,the classification results and the correct classification results are compared,the recognition accuracy and the time cosuming of the method is obtained.Finally,compared with other existing common recognition methods,the classification effect of support vector machine and BP neural network algorithm on tool wear state recognition is analyzed and compared with the results of eXtreme Gradient Boosting Tree algorithm.Therefore,we can obtain that the classification effect of the eXtreme Gradient Boosting Tree algorithm has obvious advantage.Based on the research results,the accelerated subspace iteration method and the eXtreme Gradient Boosting Tree algorithm have a good application effect on the recognition of tool wear state.This has certain practical significance for the recognition of tool wear state and the development of its monitoring system manufacturing process in the future.