Bay-Fi
This is the landing page for the Bay-Fi project, which stands for Bayesian Frequency Identification. This study introduces a novel general method that enables direct identification of the main modal frequency without requiring signal reconstruction and any prior tuning. The proposed approach employs a Bayesian framework to optimize a curve-fitting algorithm applied to highly under-sampled and randomly acquired vibration signals.
The method is particularly suited for Sub-Nyquist sampling scenarios, where conventional signal processing techniques fail due to aliasing and insufficient sampling rates. By leveraging Bayesian inference, Bay-Fi provides robust and accurate frequency estimates even from severely compressed measurements.
Here you can find more information about the project and the ongoing research activities: