Reconstructing irregularly sampled laser Doppler velocimetry signals by using artificial neural networks

The analysis of turbulent flow signals irregularly sampled by a laser Doppler velocimeter is assessed by means of ANNs. This technique has been proven to correctly predict the time evolution of turbulent signals. We are taking advantage of this ability to obtain models of unevenly sampled signals and thus be able to reconstruct and resample them at a regular pace in order to allow for their conventional analysis