Efficient computation of broadband noise propagation using Gaussian beam tracing method.

Area of Science:

• Acoustics and computational physics.
• Environmental noise modeling.

Background:

• Gaussian beam tracing is efficient for outdoor noise propagation but requires repeated computations for broadband sound.
• Existing methods are computationally intensive for broadband or multi-frequency acoustic analysis.

Purpose of the Study:

• To develop an efficient computation method for broadband noise propagation using Gaussian beam tracing.
• To reduce computational cost by decoupling ray properties from source characteristics.

Main Methods:

• The method decouples ray behaviors and source properties at high frequencies, enabling a single computation for frequency-independent ray properties.
• Ray-centered coordinates and beam dynamics are determined once, allowing simultaneous introduction of sound waves at different frequencies.

Main Results:

• Significant computational cost savings are achieved compared to multiple individual frequency computations.
• An acceleration factor of up to 35-fold was observed with 200 frequency components.
• The method was successfully applied to model noise from multiple drone propellers in urban environments.

Conclusions:

• The proposed Gaussian beam tracing method offers a computationally efficient solution for broadband noise propagation.
• This approach is valuable for analyzing complex acoustic scenarios, including urban noise and stochastic broadband problems.