A Feedforward Neural Network for Modeling of Average Pressure Frequency Response

Abstract The Helmholtz equation has been used for modeling the sound pressure field under a harmonic load. Computing fields by means of solving can quickly become unfeasible if one wants to study many different geometries ranges frequencies. We propose machine learning approach, namely feedforward dense neural network, computing average over frequency range. data are generated with finite elements, numerically response pressure, an eigenmode decomposition pressure. analyze accuracy approximation and determine how much training is needed in order reach certain predictions response.