Prediction And Optimization Of Spindle Rotation Accuracy Of Chemical Fiberwinding Machine

High speed winder is an important part of chemical fiber filament production equipment.It plays a role in stretching and winding the fiber by virtue of the rotation movement of spindle.Therefore,the rotation accuracy of spindle of winder directly affects the forming quality of silk cake.Assembly has a significant impact on the spindle rotation accuracy.Due to the large number of spindle assembly parts,complex assembly relationship and pure manual assembly,its rotation accuracy is not easy to control.How to realize the prediction of spindle rotation accuracy and optimize the existing tolerance design scheme is of great significance to guide spindle assembly.Firstly,the influence of spindle rotation accuracy on its vibration characteristics is studied through dynamic analysis;Secondly,the fusion model is established to predict the rotation accuracy;Finally,a multi-objective optimization model is constructed to optimize the geometric tolerance of spindle shaft.The research contents are as follows:（1）The influence of spindle rotation accuracy on its vibration characteristics is studied.Firstly,the spindle assembly accuracy requirements are analyzed through the spindle structure composition,and the rotation accuracy is divided into coaxiality and parallelism for follow-up research;Secondly,the spindle shaft transmission relationship is analyzed,and the spindle shaft dynamic is established according to Hamilton principle;Finally,the dynamic equation of spindle shaft is solved based on Runge Kutta method to determine the variation range of coaxiality and parallelism.（2）In order to realize the prediction of spindle rotation accuracy of winding machine,a fusion modeling and prediction method based on depth transfer learning is proposed.Firstly,the spindle tolerance is modeled based on Jacobian spinor theory,and the spindle geometric tolerance transmission dimension chain is constructed;Secondly,the precision sample set is generated for the measured assembly data of spindle to solve the problem of over fitting caused by data set;Finally,the depth transfer learning model is constructed to learn the mechanism data set and accuracy sample set,so as to realize the spindle rotation accuracy prediction,and provide evaluation indexes for the subsequent geometric tolerance optimization.（3）In order to optimize the geometric tolerance design scheme of spindle shaft of winding machine,a multi-objective tolerance optimization model driven by spindle shaft vibration characteristics is proposed.Firstly,the degree of spindle geometric tolerance is analyzed based on Monte method,and the dedicate degree ranking by the tolerance design scheme is obtained;Secondly,a multi-target tolerance optimization model considering cost and mass loss is set up,and the kinetic penalty function is draw into transform the solution process into a solution problem;Finally,the model is solved based on non dominated sorting genetic algorithm to obtain a better tolerance design scheme.（4）According to the actual engineering requirements of a winding machine manufacturing enterprise in Beijing,a micro service system is established to verify the functional modules of the above rotation accuracy prediction method and geometric tolerance optimization method,so as to provide a reliable application tool for the actual assembly guidance and tolerance design of the spindle shaft of the winding machine.To sum up,this paper realizes the effective analysis of spindle rotation accuracy,which is not only suitable for spindle rotation accuracy prediction,but also conducive to the geometric tolerance scheme design of spindle,and provides support for practical engineering application.