Research On The Health Condition Assessment Of CNC Turret Based On Condition Monitoring Data

As an important functional component of CNC machine tools,the service performance of the CNC turret directly affects the machining accuracy of the parts and the reliability of the equipment,and then affects the usability of the whole machine.Therefore,the effective assessment of the health condition of CNC turret in the production environment is helpful to carry out fault early warning and predictive maintenance,and improve production efficiency and reduce failure rate,which is very important to improve the reliability level of CNC machine tools.This is essential to improve the reliability of the CNC turret and the entire CNC machine tool.However,due to the complex structure,variable working conditions,various failure modes and other constraints,as well as the scarcity of fault sample data,the current theoretical method for the health condition assessment of the CNC turret is not perfect.In view of the frequent failures of the CNC turret rotation system,this paper mainly studies the health assessment method and application of the CNC turret rotation function system.Through the FMECA analysis of the CNC turret,the condition monitoring indicators are determined,and the monitoring methods and experimental schemes of each indicator are introduced in detail.Feature extraction,PCA dimensionality reduction and AHP feature fusion were performed on the performance indicators collected in the experiment process to obtain effective health assessment feature indicators,and the HMM was used to establish the health assessment model of the CNC turret.Finally,the real-time health monitoring and assessment system based on Spark Streaming engine was developed.The system has the characteristics of real-time collection and calculation analysis of condition monitoring data,and has certain reference value for the application of health assessment methods in the field of CNC turret engineering.The specific research content of this paper is as follows:1.In-depth analysis of the internal structure and working principle of the CNC turret,and divide the internal structure of a certain model of CNC turret into subsystems.Carry out a detailed classification of the parts under each subsystem,analyze the typical failure modes of the CNC turret,explore the failure causes of different failure modes,and judge the severity ESR and the possibility OPR of each failure mode,and finally establish failure mode,effects and criticality analysis(FMECA)of the CNC turret.2.According to the analysis results of FMECA,the possibility OPR and severity ESR are obtained,and then the comprehensive risk priority number RPN is calculated.According to the RPN value,arrange the failure modes of the CNC turret in descending order,analyze the high-frequency failure modes in the turret function system and the corresponding failure causes,finally,two continuous signals of vibration signal and current signal of the CNC turret in the process of rotation are selected as the performance indicators for the health condition assessment.Carry out a cumulative100,000 reliability operation test for a certain type of CNC turret,and arrange a signal acquisition device to build a condition monitoring platform based on Lab VIEW real-time data acquisition program to collect the rotation signal data of the CNC turret.3.Dimensional reduction processing is performed on the vibration signals of the X,Y,and Z axes and the current signals of the U,V,and W phases by the PCA to obtain the principal components of vibration and current.The influence weight of vibration signal and current signal is divided by the AHP,and the vibration signal and current signal are feature fused according to the influence weight,and finally the health evaluation index HI of the CNC turret is obtained.4.Divide the health condition space of the CNC turret into five levels of health,sub-health,normal,deterioration and failure,and use these five levels as the hidden state of the CNC turret,and the HMM is used for model training and analysis of historical condition monitoring data.According to the probability result,the current health condition of the CNC turret is judged,and the probability change trend of the health condition at the historical moment is obtained.5.Using the real-time calculation engine Spark Streaming and equipped with the health evaluation model of the CNC turret based on the HMM training,real-time calculation and assessment of the monitoring data during the rotation of the CNC turret and the assessment probability results of history are carried out.Then the probabilistic changes of health condition at all times are visualized in real time,and a real-time evaluation system for the health condition of the CNC turret is finally established.