Research On Maintenance Decision Of CNC Based On Health Status Estimation

NC machine tool is an electromechanical integration equipment that integrated bymechanical, electrical, hydraulic and computer technology. It is the carrier of realizingadvanced manufacturing technology, and it is very important in the machinerymanufacturing industry. Once NC machine tool appears accidental failures, it has had ahuge impact to the manufacturing enterprise. However, due to the complex of structure,diversity of degradation and changeful of faults, repair and maintenance of NC machinetool has been a difficult problem in engineering application.To determine the reasonable maintenance strategies of NC machine tool, this paperfirstly analysis the structure and fault mechanism of NC machine tool, divide machine toolinto four subsystems according to the function module. The key performance items aredetermined and detected periodically. Then, hidden Markov model is built that based onmultiple degradation sequences, multiple performance degradation law of NC machinetool is revealed, further the machine health status are evaluated.Then, using each subsystem’s failure probability curve which is achieved in thehealth status evaluation, calculate reliability function of the series system, and furtherdetermine the failure rate function. Next, the maximum availability model of the wholemachine is built by using the failure rate function, and the optimal preventive maintenancecycle is achieved. In addition, due to the difference in fault probability distribution ofdifferent subsystems, it may cause excess maintenance that preventive maintenance of allsubsystems at the same time. Thus put forward a minimum maintenance cost rate modelunder the condition of guarantee the maximum availability of machine. This modelintroduces a0-1matrix as decision variables, which the element represents whetherpreventive maintenance for each specific subsystem. Finally the nonlinear integerprogramming is realized by using GA algorithm. Then, the proposed method was appliedto FB260NC machine tool’s health condition assessment and maintenance decisionanalysis, the results show that the proposed method has good validity and practicability.Finally, based on the JAVA language, the prototype system of NC machine toolmaintenance decision support is developed. And the full text is summarized andprospected.