齿轮故障诊断技术的研究

本文总结了我们多年来对齿轮故障诊断的研究工作。论述了故障齿轮振动信号的数学模型;论证了齿轮故障诊断的新方法:细化复包络分析、宽带解调技术、三阶谱分析、相关谱分析和最大熵谱分析,并列举了实例。 细化复包络谱是将细化技术应用到复包络谱上,可以提高频率分辨力。理论分析与计算机模拟表明,宽带解调技术具有很高的信噪比。三阶谱与相关谱对某些故障的分析很有效。理论分析表明,对调相信号的最大熵谱估计是有偏估计,文中还给出了线性调相与线性调频两种情况下的偏移量。

Researches on Gear Fault Diagnostic Techniques

Research work on gear fault diagnostic techniques of late years in Research Institute of Vibration Engineering is presented. The mathematical model for vibration signals of de-fected gears is concisely expounded. The new methods of gear fault diagnoses, such as zoom complex envelope analysis, wide-band demodulation techniques, bispectrum analy-sis, correlation spectrum analysis, and maximum entropy spectrum analysis, are demon-strated. And illustrations listed are to show the methods to be efficient.Zoom complex envelope is formed by zoom technique applied to complex envelope in order to increase the frequency resolution. Wide-band demodulation technique is of high signal to noise ratio proved by theoretical study and computer simulation. Bispectrum and correlation spectrum are excellent for some mechanical fault forms. The theorectical analy-sis shows that the maximum entropy spectrum estimate for phase demodulated signals is biased, giving the deviations under two cases of the linear phase demodulation and the lin-ear frequency demodulation.