Research On Reliability Allocation And Prediction Technology Based On Meta Units Of CNC Machine Tools

CNC machine tools is the foundation of the country to realize industrialization and modernization,so its quality or reliability is a measure of a country's modernization and even a sign of comprehensive strength.Since the reform and opening up,higher speed,well precision,and more complex are important characteristics of home-made numerical control machine's progress,and we majored great achievements in these areas.However,compared with the developed countries,home-made numerical control machine tools has a great gap in terms of quality and reliability.CNC machine tools has the characteristics of small batch production,complex structure,and expensive,so,thronging the traditional method of statistical sampling is difficult to obtain the products' reliability index.Only the products has been put into the using process,the actual reliability can be obtained by analysis the fault information.However,it makes no sense to analysis the reliability,when the product that has been put into using process,because its reliability have been confirmed.It is very important for improving product's reliability when perform reliability allocation and prediction at the beginning of the design,so as to constraint the design process.Supported by the National Major Scientific and Technological Special Project,the reliability allocation and prediction of some machining center was studied in the process of product's development.The main contents of this paper are as follows:(1)Carefully studied the basic theory of reliability allocation and prediction,and analyzed the characteristics of CNC machine tools' reliability allocation and prediction.Machining center was taken as an example to illustrate the theory of FMA,and the divided the hierarchy structure of the product.at last “parent groups” and “offspring groups” were defined.(2)A reliability allocation method was proposed based on the interval AHP and meta action units.Through analyzing the relation between the three of “function groups”,“motion groups” and “meta action units” that come from FMA,and the characters of the groups,factors affecting reliability allocation and prediction were determined.And combining the interval analytic hierarchy process(AHP),the weights of affecting factors were given.Reliability allocation is top-down decomposition process.And considering the relation between the whole machine and Meta action units,reliability allocation mathematical model were given.The reliability of the offspring groups not only have an effect on its parental generation groups,also be affected by other groups that at the same level.Considering the effects,built a correlation matrix between the offspring groups to modify the reliability allocation method.(3)Put forward a reliability prediction method based on Meta action units and similarity theory.The reliability prediction of the numerical control is an integrated process from Meta action units to the whole product.So,firstly,by building a comprehensive evaluation cloud model to choose similar reference structures for meta action units.And then revising the reference structures' reliability data to expect Meta action units' reliability level.According to the relationship between upper and lower groups,constructed reliability prediction mathematical model which can predict the upper groups' reliability by lower groups' reliability data.Finally,machining center was taken as an example for the case analysis,and the reliability prediction value were of "groups" were given,and so as the whole machine.