2020/12/03 by Klas Pettersson, Pettersson, Klas, Andrey Karzhou +3
Engineering · #Acoustic Wave Phenomena Research #Structural Health Monitoring Techniques #Vehicle Noise and Vibration Control
paper · pdf · doi:10.48550/arxiv.2012.02276
The Helmholtz equation has been used for modelling the sound pressure field\nunder a harmonic load. Computing harmonic sound pressure fields by means of\nsolving Helmholtz equation can quickly become unfeasible if one wants to study\nmany different geometries for ranges of frequencies. We propose a machine\nlearning approach, namely a feedforward dense neural network, for computing the\naverage sound pressure over a frequency range. The data is generated with\nfinite elements, by numerically computing the response of the average sound\npressure, by an eigenmode decomposition of the pressure. We analyze the\naccuracy of the approximation and determine how much training data is needed in\norder to reach a certain accuracy in the predictions of the average pressure\nresponse.\n