Research On The Wear State Evaluation Method Of Milling Cutter Based On CSSAE

Milling cutter is the core tool in the production and processing of impeller,and its wear state evaluation in the process of processing is the guarantee of processing safety and cost saving.In this paper,the current signal of the spindle motor of CNC machine tool is used as the monitoring signal to design the data acquisition experiment,and an experimental platform is built to obtain the data set of milling cutter wear signal.The data set is verified and tested by the deep learning algorithm to determine the suitability The algorithm model for evaluating the wear state of milling cutter is developed,and the data flow transfer and online monitoring system are realized by combining the big data platform and LabVIEW programming.The main research contents of this project are as follows::(1)The literature at home and abroad is analyzed synthetically and the mechanism of tool wear is studied.Through the theoretical analysis and formula derivation,the feasibility of using the current signal of machine tool spindle instead of the cutting force signal as the basis for the analysis and evaluation of the wear state of milling cutter is verified.The corresponding data acquisition test scheme is designed to collect the data sets of round nose milling cutter and ball end milling cutter,and the experimental data set is preliminarily analyzed and processing.(2)Based on CSSAE,the wear state evaluation model of milling cutter is constructed.Combined with CNN's strong data compression ability and SSAE's high test accuracy,CSSAE algorithm model is constructed,and parameters are adjusted to determine the optimal network parameters by using data set validation,so that it has high test accuracy and sample training speed is effectively improved,which can be used as an algorithm model to evaluate the wear state of milling cutter.(3)Combined with Hadoop and spark,the Hadoop spark big data framework is constructed.The collected experimental data is stored in HDFS in the form of data sets for calling.Then,CSSAE algorithm program is used to call the datasets and analyze them in Spark to verify the feasibility of big data platform analysis.(4)Based on the advantage that LabVIEW programming can embed compatible data acquisition,storage and analysis algorithm,an on-line evaluation system of milling cutter wear state is developed.The wear state of milling cutter is analyzed by CSSAE algorithm model,and online monitoring is realized.The wear curve is fitted by RNN to predict the wear trend of milling cutter.