基于DCT和FFT谱熵分析的机械传动系统状态检测与趋势预测

Analysis of Spectral Entropy Based on DCT and FFT and Its Application to the Condition Detection and Tendency Estimate of Mechanical Transmission System

采用DCT和FFT两种方法分别对同一组机械传动齿轮振动信号进行了时域到频域的变换,应用谱熵分析的方法对变换后的信号进行研究和分析处理,计算了不同转速下不同故障类型的齿轮系统信号谱熵值,给出了不同转速下的二维谱熵分布图,并且对系统的工作状态进行了区别和趋势分析预测。结果表明:DCT变换能更好地将时域信号的特征在频域中集中表现出来,可以更好地将不同状态的机械传动齿轮信号加以区分、分离与聚类;FFT变换能够对机械传动齿轮的工作状态和恶化工作趋势进行分析、估计和预测。两种方法的应用为机械传动及其机械系统的工作状态分析、监控和诊断提供了可靠的手段。

This paper studied and analyzed the vibration signals of gear of a mechanical transmission based on the spectral entropy by using DCT and FFT, the working condition was detected and runing tendency was estimated for the transmission system. The results show that DCT is good at exhibiting the signal features in frequency-domain and is able to distinguish and cluster the different gear signals for mechanical transmission while FFT is clever at analyzing and forecasting the deterioration tendency for gear of mechanical transmission. It provides an dependable approach to detect working condition and diagnose runing tendency for mechanical transmission and mechanical systems.