Recognition Of The Degradation State Of The Preload Force Of The Ball Screw Pair Of The Vertical Machining Center Feed Axis System

The ball screw is the key rolling component of the CNC machine tool.Its performance is an important factor affecting the service life and reliability of the CNC machine tool.Since the preload directly affects the load distribution,rigidity,friction and wear and fatigue life of the ball screw pair,the preload is an important index to evaluate the service performance of the ball screw.Based on the National Natural Science Foundation of China and the National Science and Technology Major Project,this paper conducts research on the degradation state of the preload of the ball screw.First,the theoretical model of the degradation of the preload is established,and the preload and vibration signal acquisition system of the ball screw is built to collect the preload and vibration signals.Then,starting from the model-driven and data-driven respectively,a recognition model of the preload of the ball screw is constructed based on the characteristic frequency and the preload of the ball screw and the entropy weight method?KPCA?GS?SVM Force state recognition model.Two recognition models on the test bench and vertical machining center are verified,compared,and analyzed.The details are as follows:(1)Based on the Hertz contact theory and Archard wear theory,the degradation mechanism of the preload of the ball screw is analyzed.the preload degradation theoretical model is established,and the division of the degradation state of the preload of the ball screw and the calculation of the preload are completed.Combining the degradation model of the preload force,the dynamic model of the ball screw feed transmission system is established,the dynamic equations are established and solved to obtain the theoretical value of the characteristic frequency corresponding to the preload in different states.The characteristics are calculated corresponding to the critical preload frequency to establish the recognition model of the preload state of the ball screw,based on which the recognition of the degraded state of the preload force is realized in the ball screw feed drive system.(2)According to the degraded characteristics of the preload of the ball screw being fast and then slow,and the inconvenience of preload acquisition in the feed system,the preload and vibration signal acquisition schemes based on the test of the bench and the vertical machining center were designed respectively.The collection of preload and vibration signal.Completed the extraction of the characteristic parameters of the vibration signal from the three aspects of the time domain,frequency domain and time-frequency domain,constructed a hybrid domain feature set,and realized the optimization and dimensionality reduction of the feature set using the entropy method and KPCA.The influence of the three different parameter optimization algorithms GA,PSO,GS of SVM on the recognition effect is analyzed,and finally the recognition model of the preload state of the ball screw based on the entropy weight method?KPCA?GS?SVM is established.(3)On the test bench and the vertical machining center respectively,we verify the preload state recognition model of the ball screw based on the characteristic frequency and the preload state recognition model based on the entropy weight method?KPCA?GS?SVM.The two models are compared and analyzed in three aspects: recognition accuracy,data reliability,and difficulty of implementation,and their respective application strategies are proposed.