Parameter identification of a physical model of brass instruments by constrained continuation

Numerical continuation using the Asymptotic Numerical Method (ANM), together with the Harmonic Balance Method (HBM), makes it possible to follow the periodic solutions of non-linear dynamical systems such as physical models of wind instruments. This has been recently applied to practical problems su...

Full description

Bibliographic Details
Main Authors: Fréour Vincent, Guillot Louis, Masuda Hideyuki, Vergez Christophe, Cochelin Bruno
Format: Article
Language:English
Published: EDP Sciences 2022-01-01
Series:Acta Acustica
Subjects:
Online Access:https://acta-acustica.edpsciences.org/articles/aacus/full_html/2022/01/aacus210088/aacus210088.html
Description
Summary:Numerical continuation using the Asymptotic Numerical Method (ANM), together with the Harmonic Balance Method (HBM), makes it possible to follow the periodic solutions of non-linear dynamical systems such as physical models of wind instruments. This has been recently applied to practical problems such as the categorization of musical instruments from the calculated bifurcation diagrams [V. Fréour et al. Journal of the Acoustical Society of America 148 (2020) https://doi.org/10.1121/10.0001603]. Nevertheless, one problem often encountered concerns the uncertainty on some parameters of the model (reed parameters in particular), the values of which are set almost arbitrarily because they are too difficult to measure experimentally. In this work we propose a novel approach where constraints, defined from experimental measurements, are added to the system. This operation allows uncertain parameters of the model to be relaxed and the continuation of the periodic solution with constraints to be performed. It is thus possible to quantify the variations of the relaxed parameters along the solution branch. The application of this technique to a physical model of a trumpet is presented in this paper, with constraints derived from experimental measurements on a trumpet player.
ISSN:2681-4617