滚珠丝杠导程误差分析与计算机辅助诊断研究

【作者】 王植；

【导师】 殷爱华；

【作者基本信息】 南京理工大学 ， 机械电子工程， 2004， 硕士

【摘要】 本文所研究的内容是汉江机床厂滚珠丝杠磨床改造项目的一部分，介绍了一种利用在线测量获得的滚珠丝杠导程误差对丝杠磨床进行在线工况监测和诊断的方法。 为了分析滚珠丝杠的导程误差，采用了多种方法包括：时域的回归分析、直方图分析、均值—标准偏差控制图分析；频域的功率谱密度分析、倒频谱分析和小波包分析。由于滚珠丝杠导程误差的复杂性，不能直接对它进行功率谱密度分析和倒频谱分析，必须先对它们进行小波包频带分解。这种方法证明是很有效的。 所有的程序都使用Matlab语言编写的。Matlab语言具有非常强大的信号处理功能。最后在对误差数据分析实验取得的参数的基础上，利用Matlab语言编了一个操作界面，从而能够很方便的使用。更多还原

【Abstract】 What the paper studying is a part of the project of reformation of ball screw grinding machine tool of Hanjiang Machine Tools Factory. This paper introduces a on line working status monitoring and assistant failure diagnose system of ball screw grinding machine using ball screw lead error data which be acquired just after the ball screw is grinded on the machine.Several methods have been used to analysis the error signal in the paper, include regression analysis, histogram analysis, mean- standard deviation control chart ,they are analysis methods in time domain ;power spectrum density analysis, cepstrum analysis and wavelet package technology, they are analysis methods in frequency domain. Because the ball screw lead error data is so complex that we can’t do the power spectrum density analysis and cepstrum analysis directly. Before doing this we must use wavelet package frequency band decomposing technology to decompose the frequency band of the screw lead error data . These methods have been proved to be very useful.All the programs were written in Matlab language. The Matlab language has very powerful ability in signal processing. In order to be used conveniently ,the software have a friendly interface as we can see in the paper, which is bsed on the parameters acquired from the experiments on the analyse of lead error data.更多还原