Reliable Running Time Estimation And Failure Prediction Of Motorized Spindle

CNC machine tools are the foundation of modern equipment manufacturing,and motorized spindles are the core components of CNC machine tools,and their performance and health status directly affect the overall operating status of CNC machine tools.Once the motorized spindle fails,it will cause problems such as shutdown of the CNC machine tool for maintenance,which will affect the quality and production efficiency of the product.Aiming at the characteristic of the performance degradation of the motorized spindle in the early stage of operation,the reliability of the Motorized Spindle ’s reliable running time and the prediction of hidden troubles are proposed.During the operation of the motorized spindle,its characteristic parameters are monitored in real time,so that it can predict the time node of possible failure in the future when the failure has not occurred;and the failure traceability of the factors that may cause the failure.Preventive maintenance can be achieved through the reliable running time estimation of the motorized spindle and the prediction of hidden troubles of the fault,which can avoid the failure of the motorized spindle to the greatest extent,extend the life of the CNC machine tool,and have the important meaning of reducing the maintenance cost and improving the stability of the equipment.The main contents of this article are as follows:(1)Establish the performance degradation model of the Motorized Spindle.First,according to the system configuration of the Motorized Spindle,the bearing vibration and stator current are used as characteristic parameters to establish the performance degradation model of the mechanical and electrical parts;then the performance degradation model of the Motorized Spindle is obtained through model fusion.(2)Using Bayesian method to estimate reliable working hours.First,a polynomial is used to perform curve fitting using the least squares method,and the optimal fitting model is determined by the value of the penalty function,and the life cycle curve of the Motorized Spindle is obtained.Then predict the time node at which the failure may occur in the future at the current moment,and obtain the reliable running time.Then use Bayesian method to predict the probability distribution of model parameters.Obtain the probability distribution of the model parameters,make the reliable running time more reliable,realize the predictive maintenance strategy,avoid the failure of the motorized spindle to the greatest extent,and prolong the service life of the CNC machine tool.(3)Use the method of model parameter change rate to predict the hidden trouble of failure.First,according to the performance degradation model of the mechanical part and the electrical part of the Motorized Spindle,the change rate of the model parameters of the two parts is obtained,and then the probability of failure of the motorized spindle is evaluated according to the change rate of the model parameters,and the fault is traced and solved To predict the problem of hidden troubles,preventive maintenance should be carried out as early as possible,so that the motorized spindle can work more effectively.Extend the operating life of the motorized spindle.To sum up: Aiming at the characteristics of implicit degradation of the performance degradation of the motorized spindle in the initial stage of operation,a study on the estimation of the reliable running time of the Motorized Spindle and the prediction of the hidden troubles of the fault is proposed.Through the established performance degradation model of the Motorized Spindle,reliable running time can be obtained and the hidden troubles can be traced back.The validity of the method is proved by experiment simulation.