Development Of Economical Tool Wear Condition Monitoring System

During the process of machining parts,tool is indispensable.And tool condition monitoring is also a crucial segment.In view of the economic tool condition monitoring system,this paper proposes a method based on motor current signal to monitor the tool condition in the machining process.On the one hand,it is necessary to prove the correlation and replaceability between motor current signal and cutting force signal.On the other hand,it is necessary to prove the effectiveness and practicability of motor current signal applied to tool condition monitoring.The specific contents are as follows:(1)Introduce the forms and causes of tool wear,and divide the process of tool wear into three stages according to the standard of tool blunt.Carry out the tool wear monitoring test,and analyze the motor current signal and cutting force signal through time domain,frequency domain and time-frequency domain,in order to demonstrate the effectiveness of them.Meanwhile,it was preliminarily observed that there was a certain similarity between the motor current signal and cutting force signal.(2)Carry out the tool wear monitoring experiment based on motor current signal.The three-phase current signals of spindle motor and feed motor of two milling machines with different production dates are collected at different sampling frequencies to analyze their response characteristics,and the variation rule of motor current signal in the process of machining is obtained.At the same time,the mathematical model between the motor current signal and the cutting force is established,and the linear correlation between the motor current signal and the cutting force signal is proved by Pearson correlation analysis.It is proved that the correlation between the two signals is high in the process of NC machining,and the motor current signal can be used to replace the cutting force signal for tool condition identification.(3)Extract the time domain features of motor current signal,and take advantage of the mutual information coefficient method for feature selection.Five features are optimized,including root mean square(RMS),mean,peak,variance and kurtosis index.Based on these five optimal time-domain features,the fusion algorithm of minimum feature divergence weight is used to construct the health indicator,which can obviously observe the tool breakage and edge collapse,and roughly classify the stage of tool wear.Based on the motor current signal,the RNN recurrent neural network is used to recognize the condition of machining tools.In addition,the accuracy,recall and F1 score are introduced to evaluate the performance of the model.The results show that the model is effective and the motor current is effective for tool wear condition identification.(4)For the purpose of practicability of economical tool condition monitoring system,design the hardware composition of the monitoring system,and use the GUI toolkit wx Python to design the system interfaces of different functional modules,including system login interface,data acquisition and analysis interface and intelligent condition identification interface.The economical tool condition monitoring system can make users know about the basic information and the wear stage of the tool clearly and intuitively,so it possesses practicability.