Identification of a Time-varying Beam Using Hilbert Vibration Decomposition

The present work is concerned by modal identification of time-varying systems. For this purpose, a method based on instantaneous frequency identification and synchronous demodulation is used to extract modal components from recorded signals. The proposed method of iterated sifting process is based on the Hilbert Vibration Decomposition (HVD) technique which is used to extract the instantaneous dominant vibrating component at each iteration. A source separation preprocessing step is introduced to treat multiple degree-of-freedom systems in an optimal way. Sources are used as reference signals to get a single instantaneous frequency of each mode for the demodulation on all the channels. The algorithm is presented and is applied to numerical simulation of a randomly excited time-varying structure for illustration purpose. The investigated structure is made up of a beam on which a non-negligible mass is traveling. The variable location of the mass results in changes in modal parameters.