Research On Reliability Analysis Method Of CNC Milling Machine Based On Multi-source Data Fusion

With the continuous development of artificial intelligence technology and the continuous integration with manufacturing technology,the manufacturing industry is in a new stage of transformation to intelligence,and intelligent manufacturing technology has become an inevitable trend in the future development of manufacturing.Compared with traditional machine tools,CNC machine tools have the advantages of high efficiency,high precision,multi-function and high degree of automation,and have been widely used in aerospace,military equipment,energy equipment and other fields.However,high-grade CNC machine tools work in a complex and changing environment with demanding processing requirements,leading to the inevitable degradation of machine performance,which seriously affects the productivity of the machine and the quality of processed products.The reliability level of CNC machine tools has become one of the important indicators of equipment manufacturing capability.With the development of signal processing,artificial intelligence and other technologies,data collection and reliability assessment of CNC machine tools in service through artificial intelligence methods have become an important way of equipment management and maintenance.And how to assess the reliability of equipment from diverse sensor data has become a new problem facing the operational reliability assessment of CNC machine tools.This thesis aims at the smooth and reliable operation of CNC milling machines and conducts research on the reliability of CNC milling machines,with the following main contents.(1)Fault tree analysis is carried out for CNC milling machines to determine their weak subsystems and weak links.Taking CNC milling machine as the object,the structural characteristics and functional principles of the CNC milling machine are analyzed,and the CNC milling machine is divided into six subsystems,and the subsystem fault tree of the CNC milling machine is established.Then,by carrying out qualitative analysis and quantitative calculation of the fault tree,the minimum cut set of CNC milling machine is find out and the reliability of CNC milling machine and the importance of events at the bottom of the fault tree is calculated to determine the weak subsystems and weak components of the CNC milling machine.(2)To address the performance degradation of CNC milling machines,an operational reliability analysis method based on current signal fusion is proposed.The method uses multiple current signals and builds multiple SAE models to extract the deep degradation features embedded in each signal.The extracted features are filtered and then the operational reliability assessment model of the CNC milling machine is established through the state distance-based reliability calculation method to assess the operational reliability of the equipment.Finally,the validity and feasibility of the method are verified through examples.(3)To address the problem that a single sensor signal source cannot accurately and comprehensively characterise the degradation state of the equipment,this thesis proposes a method to analyse the operational reliability of CNC milling machines based on multi-source signal fusion.The method first performs data pre-processing on multi-sensor data to achieve data-level fusion,and then uses them as input data to extract features in the multi-source signals using multi-channel 1D-CNN,and establishes an equipment operational reliability assessment model based on the state distance-based operational reliability calculation method to evaluate the equipment operational reliability.Finally,the validity and feasibility of the method are verified through examples.