Neural network data analysis for laser-induced thermal acoustics

A general, analytical closed-form solution for laser-induced thermal acoustic (LITA) velocimetry signals using homodyne or heterodyne detection, and electrostrictive and thermal gratings, has been derived. A one-hidden-layer feedforward neural network was trained using back-propagation learning and a steepest descent learning rule to extract the speed of sound and flow velocity from a heterodyne LITA signal. The accuracy was determined with a second set of LITA signals that were not used during the training phase. The accuracy is better than that of a conventional frequency decomposition technique while being computationally as efficient. (Original abstract - amended)