Research On The Status Monitoring And Health Evaluation Method Of Typical CNC Machine Tools

With the global manufacturing industry develops towards smart manufacturing,computer numerical control(CNC)machine tools have become the core manufacturing equipment.However,the frequent failure of machine tools in the manufacturing process seriously restricts the efficient and high-quality of the manufacturing.In order to monitor the state of machine tools and avoid the occurrence of failures as much as possible,this paper designs a state monitoring system for two typical CNC machine tools: a vertical machining center and a CNC horizontal lathe.The system monitors the status of CNC machine tools by collecting and analyzing the status data of the key components of machine tools.The health status of the machine tools is evaluated so that the predictive maintenance and reduce of the machine tool failure rate are achieved.The main contents of this paper are as follows:(1)Statistical analysis of the failures of CNC machine tools are carried out.The key mechanical components of CNC machine tools are determined through the failure mode analysis method.The vibration and temperature are selected to realize the status monitor of the key mechanical components according to the failure performance mode.The external sensor and Ethernet-based data collection method are determined through the analysis of the vibration signal,temperature signal and the failure mechanism of the CNC system,to achieve accurate and efficient collection of the status data.(2)The hardware and software structure of a data acquisition and monitoring system are designed based on the functional requirements.The design of hardware mainly includes the hardware platform construction and the optimization of the installation location layout of sensors.The software is a graphical interface application program developed based on the C ++ language and MATLAB,which realizes the functions of data collection,status data monitoring and data analysis.(3)The vibration data collected by the system are pre-processed by eliminating trend terms and reducing the wavelet threshold noises.And then the feature values of the vibration data are extracted in time domain,frequency domain,and time-frequency domain.Finally,the health status of the machine tools are evaluated by analyzing the most sensitive characteristic data in three kinds of experiments of treadmill,light load cutting and heavy load cutting.(4)A health evaluation method of CNC machine tools is established based on fuzzy gray clustering method and combined weighting method.Firstly,the health status of CNC machine tools is divided into four categories: health,sub-health,available and failure.Then,the health evaluation of key components is conducted via gray clustering method.Afterwards,the health evaluation of the whole machine tool is finished via fuzzy judgement based on the health status of the key components.Finally,the neural network method is used to quickly evaluate and verify the evaluation results.The results obtained by the neural network method are consistent with the evaluation results of the fuzzy gray clustering method and the combined weighting method,which further proves the accuracy and usability of the proposed evaluation method.