The main purpose of this paper is to represent a methodenabling vibration components to be extracted from a high-resolution Short-Time Fourier-Transformation (STFT) based spectrogram assessed as an image to support transient analysis on rotating machines. Therefore, an improved STFT algorithm was developed to allocate and utilize computational memory more efficiently. The resulting spectrogram was compressed into a grey-scale image without any kind of information loss and was used for further image processing methods to obtain details about the vibration components. Furthermore, differential and moving average predictive tracking algorithms were developed for frequency ridge evaluation in the spectrogram image. For further analysis, the obtained results were transformed back with an inverse transformation method from image-space to time–frequency plane. Moreover, these results are able to be used to estimate the speed of rotation of the machine and to observe the frequency components. The methods were tested and validated with simulated signals and transient measurements on rotating machines. With the combination of vibration signal- and image processing techniques the evaluation time and computational resource requirements are decreased enhancing more efficient and accurate analysis, nevertheless opens the possibility of a real-time condition monitoring based on a basic vibration measurement.

本文的主要目的是代表一种方法，该方法能够将振动分量从基于图像的高分辨率短时傅立叶变换（STFT）频谱图中提取出来，以作为图像来支持旋转机器上的瞬态分析。因此，开发了一种改进的STFT算法以更有效地分配和利用计算内存。所得频谱图被压缩为灰度图像，而没有任何类型的信息损失，并且被用于进一步的图像处理方法以获得有关振动分量的详细信息。此外，还开发了差分和移动平均预测跟踪算法，用于频谱图图像中的频率脊评估。为了进一步分析，用逆变换方法将获得的结果从像空间变换到时频平面。此外，这些结果能够用于估计电机的旋转速度并观察频率分量。通过模拟信号和旋转机器上的瞬态测量对方法进行了测试和验证。结合振动信号和图像处理技术，减少了评估时间和计算资源需求，从而提高了分析的效率和准确度，但为基于基本振动测量的实时状态监视提供了可能性。通过模拟信号和旋转机器上的瞬态测量对方法进行了测试和验证。结合振动信号和图像处理技术，减少了评估时间和计算资源需求，从而提高了分析的效率和准确性，尽管如此，它仍然提供了基于基本振动测量进行实时状态监测的可能性。通过模拟信号和旋转机器上的瞬态测量对方法进行了测试和验证。结合振动信号和图像处理技术，减少了评估时间和计算资源需求，从而提高了分析的效率和准确度，但为基于基本振动测量的实时状态监视提供了可能性。