A feedforward neural network for modelling of average pressure frequency response

TL;DR

This paper introduces a neural network model to efficiently predict the average sound pressure response over frequency ranges, reducing computational costs compared to traditional Helmholtz equation solutions.

Contribution

It presents a novel machine learning approach using a feedforward neural network to model average pressure frequency response, trained on finite element data.

Findings

Neural network accurately predicts average pressure response.

Training data size impacts prediction accuracy.

Method reduces computational effort for frequency response analysis.

Abstract

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