Research On Evaluation Method Of Running Reliability For CNC Equipment

CNC equipments are widely used in modern industry as mechanicalmanufacturing industry machine tools, such important equipment once failure, notonly leads to the loss of the equipment itself also will bring a host of other greateconomic losses. CNC equipment is different from the general mechanical equipment,because of its organization, function, working environment is complex, due toequipment failure mode diversity, so it is important to forecast and evaluation of thereliability of its operation process. The classical reliability assessment method basedon failure data to carry out has stringent requirements to the data information of thesamples, it does not apply to the single CNC equipment and instantaneous reliabilityevaluation, so it is necessary to put forward another methods for evaluating thereliability. This paper funded by the national natural science funds, using degradationdata of numerical control equipment, has effective completed the prediction andevaluation of CNC equipment operational reliability based on grey time series model,hidden semi Markov model. The main work is as follows:(1) Introduced the theory of performance degradation, analyzed the performancedegradation phenomenon of CNC equipment during operation, designed a plan tocollect the performance degradation data reflects the real-time status of CNCequipment, extracted the characteristic parameters of external representation ofperformance degradation of CNC equipment. A new method was put forward toestablish the model of CNC equipment performance degradation based on grey timeseries model, showed the grey prediction model GM (1,1) and time series AR modelmodeling process, through the use of models to understand the degradation ofequipment performance, to predict the time of equipment failure.(2) Elaborated the basic concepts and definitions of Markov chain, Markov andhidden Markov model (HMM), Attention and improvement should be paid to in modeltraining process as well as the problems in application. Given the basic steps ofoperational reliability evaluation method based on HMM, CNC equipment reliabilityevaluation method based on HMM described the relationship between external performance parameters and internal state, combined with the example calculates theinstantaneous reliability in the operation of CNC equipment, understand reliabilitychanges during equipment works.(3) Introduced an extended form is hidden semi Markov model (HSMM) based onthe improvement of HMM, first introduced the parameters in HSMM and thesignificance of the generalized algorithm, secondly, through the introduction of fastrecursive algorithm to solve the computation and storage of exponentialmultiplication problems in the training process, by using the calibration technique toavoid the data underflow problems in model training, and improved the robustness ofmodel training. Finally, the reliability of CNC equipment was evaluated usingmodified HSMM, compared and analyzed between HMM and HSMM.