Research On Condition Monitoring And Typical Fault Early Warning System Of CNC Lathe

With the implementation of "Made in China 2025" strategy,the equipment manufacturing industry,as the basic industry for national economic development,is developing towards the trend of intelligence and digitalization.CNC machine tools have always had the title of "industrial machine",is widely used in automotive,aerospace,shipping and other fields.As the core equipment of manufacturing industry,it has greatly improved the production efficiency and quality.In recent years,with the continuous emergence of "intelligent factory" and "unmanned factory",once the failure of CNC machine tools,not only affect the production efficiency,seriously even lead to the shutdown of equipment,causing incalculable losses.In order to ensure the safe and healthy operation of CNC machine tools for a long time,it is particularly important to monitor the operation status of CNC machine tools and give early warning of faults.In order to solve the above problems,this paper takes the widely used CNC lathe as the research object,on the basis of studying the structure,monitoring parts,monitoring signal selection,signal analysis and processing,condition monitoring and fault warning algorithm of CNC lathe,proposes a complete set of CNC lathe condition monitoring and typical fault warning method,and adopts MATLAB and LabVIEW hybrid programming method to develop the CNC lathe condition monitoring and typical fault warning system,and finally carries out relevant experiments to verify the effectiveness of this paper's condition monitoring,fault warning algorithm and system.The main contents are as follows:(1)condition monitoring method analysis of CNC lathe.By analyzing the structure of CNC lathe,the key monitoring parts of CNC lathe are defined,vibration and current signals are selected as characteristic signals to represent the condition of CNC lathe,the condition of key parts of CNC lathe is characterized by vibration signals,and the condition of transmission system is characterized by current signals.On this basis,the information acquisition system of CNC lathe is built by using external vibration sensors,built-in current sensors,data acquisition cards and other hardware.(2)research on state monitoring model of CNC lathe.Through analysis of signal processing,the specific principle of eigenvalue extraction,condition monitoring algorithm research,put forward a set of complete condition monitoring model,mainly including singular value decomposition(SVD)and time domain analysis,frequency domain analysis and time-frequency domain analysis of signal analysis and processing,eigenvalue extraction algorithm,The condition monitoring algorithm of key parts of CNC lathe based on BP neural network and the condition monitoring algorithm of CNC lathe drive system based on cosine similarity are proposed to realize the state monitoring of CNC lathe.(3)research on typical fault early warning model.A typical fault warning model is established by studying the specific principle of fault warning algorithm.mainly including the selection of early warning components and indicators,the phase space reconstruction technology,the support vector regression machine(SVR)early warning algorithm,Chebyshev inequality algorithm,etc.,the fault early warning threshold is determined,and the typical fault of CNC lathe--rolling bearing fault is realized.(4)condition monitoring of CNC lathe and development and verification of typical fault warning system.The overall framework of the system and the main function modules are described in detail,and the use of MATLAB and LabVIEW hybrid programming integrated the previous main algorithm to develop the system function module,formatting a complete CNC lathe state monitoring and typical fault warning system.Finally,the experiment of CJK6132 CNC lathe is carried out to verify the validity of the CNC lathe information acquisition system and the lathe condition monitoring model.The effectiveness of the fault warning model is verified by relevant simulation experiments.The condition monitoring of CNC lathe and the function test of typical fault warning system are completed,which shows the effectiveness of the model and system.Through the research of this paper,the condition monitoring and fault warning model of CNC lathe and the developed system have good practicability,which has certain theoretical and engineering application value for the long-term safe and healthy operation of the machine tool.