基于振动信号分析的滚珠丝杠副预紧力监测

Monitoring of Preload of Ball Screw Based on Vibration Signal Analysis

滚珠丝杠副预紧力衰退将导致滚珠丝杠副的性能的下降,因此提出一种预紧力状态识别模型实现对预紧力状态的监测,分析了利用振动信号对预紧力变化趋势进行跟踪监测的可行性。通过试验模拟了预紧力衰退的三个阶段,采集各阶段的振动信号,利用小波包分解对不同频带上的能量信号进行提取;然后将信号能量和时域信号中部分统计量作为特征向量导入至SVM中进行训练。测试结果表明,小波包与支持向量机结合的分辨准确率达到90.84%。研究结果可用于对滚珠丝杠副预紧力衰减变化情况的跟踪,对于促进滚珠丝杠副健康状态监测方法的发展具有重大的意义。

The decline of the pre-tightening force of the ball screw pair will lead to the decline of the performance of the ball screw pair.Therefore,this paper proposes a pre-tightening force state recognition model to realize the monitoring of the pre-tightening force state.This paper analyzes the feasibility of using vibration signal to track and monitor the change trend of preload.This paper elaborates the three stages of ball screw's preload decay through experiments,collecting each stage's vibration signal.Use WPD to extract energy signals in different frequency bands;then the signal energy and some statistics in the time domain signal are imported as feature vectors to SVM for training.According to the testing results,the resolution accuracy of the combination of WPD and SVM reaches 90.84%.The research results can be used to track the change of pretension decline in the running of ball screw pair.,and are of great significance for promoting the development of the method of monitoring the health of the ball screw pair.