Research On Fault Trace And Analysis Technology Of NC Machine Tool Based On Meta Action

The complex numerical control machine tool is one of the most important parts of modern industry,once the complex numerical control machine tool breaking down,less lead to production interruption and even endanger personal safety.So it is necessary to reduce the fault as much as possible and ensure the safe and reliable operation of numerical control equipment.Because of the system of the mechanical and electrical products,such as CNC machine tools,has complexity and strong coupling.During the process of product running to failure,the state change of the product presents the characteristics of fuzzy uncertainty.Ths feature is the obstacle of diagnosis,has brought considerable difficulty to the accurate diagnosis of the fault of CNC machine tool.Based on the above background,this paper supported by the project of the National Natural Science Foundation of China,an approach to fault tracing and analysis of NC machine tools from the point of view of action was proposed,The research method of fault tracing and analysis includes the following parts.First,this paper analyzes the purpose and significance of the research on fault tracing technology of CNC machine tools,then introduces the current research status of fault tracing technology,and expounds the structure frame of the paper.Second,bayesian network diagnosis model based on meta-action was proposed.Considering the complexity of the internal structure of NC machine tools,this chapter combines the characteristics of motion and power transmission in the realization of NC equipment functions by meta-actions.Firstly,the function-motion-action decomposition method(FMA)was used to decompose the whole machine to get the meta-action,and then the Bayesian network has the advantage of analyzing and solving the fuzzy uncertain things.A diagnosis network model with meta-action and fault phenomena as nodes was constructed to locate the fault-causing meta-action elements.The fault meta-action unit located by diagnosis in this chapter is the foundation of follow-up analysis and research.Third,recognition of key meta action units based on fuzzy PageRank method.The key meta action unit is the meta action unit which has the greatest influence on the state of the fault element action unit.In order to analyze the potential fault hidden trouble,fuzzy PageRank algorithm was proposed to determine the key meta action unit.In this paper,the triangle fuzzy number was used to represent the fuzzy influence relation between the relative meta-action units from the angle of the structural relationship between the meta-action units,and based on this,the adjacency relation matrix was constructed.Then the PageRank algorithm was used to calculate the degree of influence between the two meta-action units which were interrelated.Combined with the failure probability of the meta action unit,and the state influence degree between the element action units was determined.According to the state influence degree between the correlation element action units,the meta action unit which has the greatest influence on the state of the fault element action unit was determined as the key meta action unit for subsequent analysis.Fourth,fault tree analysis based on meta-action unit.In fault tree analysis,each minimum cut set was one of the causes of the top event.It was of great significance to determine the correlation between the top event and the minimum cut set,which is of great significance to the diagnosis and location of the fault.In order to reduce the difficulty of modeling,a fault tree analysis model based on meta-action unit was established in this chapter.The fault causes were analyzed in the meta-action unit,and the minimum cut set of the meta-action unit fault tree was obtained.Then the probability of bottom event was described by trigonometric fuzzy number,and the fuzzy probability of top event was calculated by combining the fault tree structure of meta action unit,and then the bottom event was determined.Finally,the minimum cut set,which was the most relevant to the top event,was determined by the weighted improved grey correlation model,in order to realize the accurate location of the fault cause.