Research On Monitoring And Fault Diagnosing Technologies Of Manufacturing Equipments For Agile Maintenance

Along with the rapid development of computer, communication and automation technologies, manufacturing equipment is having improving automation level, more complex structures and functions, and increasing potential fault possibilities and modes. Further more, some fault may cause chain reaction, and the whole system fail. So it is crucial to construct a full-functioned monitoring and fault diagnosing system(MFDS), in order to realize condition monitoring and fault diagnosing, and maintain the equipment works normally and reliably. In the sight of agile maintenance (AM), this paper takes manufacturing equipment as study object, to research the MFDS's integrated application technologies.Based on the analysis of manufacturing equipments'monitoring and diagnosing requirements and application status, the description model, framework model and function model of MFDS are presented. Especially, it is given that the MFDS is composed of field monitoring centre and remote fault diagnosing centre.On the basis of deeply research on OPC specifications, wireless communication and video transmission technologies, a wireless video monitoring and control solution based on OPC for manufacturing equipments is brought forward. The key technologies of video display, equipment control with PDA based on wireless network are implemented, which makes PDA to be a convenient tool in MFDS.Then, based on the study of Bayesian Network(BN) theory, the modeling method and flow of constructing BN based on expert knowledge is brought forward. Also, with the study of Probability Propagation in Trees of Cluster(PPTC) algorithm, a new diagnosing decision making method based on possibility-cost rate is given.Finally, a MFDS prototype system is applied to an automatic assembly system, the application shows the MFDS platform aimed for AM is feasible, and the PDA used as equipment monitoring and control client based on wireless network can meet the requirement of real-time communication and reliability. It also shows the validity of modeling method and inference platform based on BN.