Partial sound source estimation with Helmholtz inverse beamforming as a part of pass-by noise virtualization

Abstract

Noise has an impact on human health. Therefore, pass-by noise of motor vehicles is legally regulated. The latest norm of the Economic Commission for Europe is ECE R51.03. According to this norm, pass-by noise of motor vehicles must be reduced to a sound pressure level of less or equal 68 dB(A). This poses a challenge for original equipment manufacturers. Firstly, they need a detailed analysis on the influence of contributing vehicle components such as engine, exhaust system and tires to the over-all pass-by noise. Secondly, original equipment manufacturers strive to shorten development cycles and lower production costs. Therefore, predictions of the expected physical behaviour of future cars have gained increasing importance. This leads to the emergence of new technological concepts like digital twins. This paper presents an extension of our latest approach to partial sound source analysis of simulated pass-by noise, i.e. Helmholtz Inverse Beamforming. Moreover, we present the embedding of the results of Helmholtz Inverse Beamforming in a comprehensive virtualization concept of pass-by noise engineering. By combining Helmholtz Inverse Beamforming with machine learning predictions of future cars, this concept enables us to derive target values of acoustic components regarding their pass-by noise.