Two levels of Bayesian Dissipation Analysis in impedance tube measurements

Accurate impedance tube measurements are often compromised by additional absorption due to air dissipation, particularly at high frequencies. Various effects contribute to air dissipation in a tube, and numerous models have been developed to describe the dissipation. This work explores different dissipation models and estimates the dissipation accurately using Bayesian inference. The exploration involves two levels of Bayesian inference: the higher level of inference addresses the model selection and the lower level of inference estimates critical parameters in models, such as sound speed and dissipation factor. The experimental data are derived from impedance tube measurements of the empty tube, and the comparison is extended to high frequency. Based on the analytical results and experimental validation, Bayesian inference is able to estimate and select dissipation models for the impedance tube measurements at high frequencies.