Differential Volterra filter: A two-stage decoupling method for audible sounds generated by parametric array...

Area of Science:

• Acoustics
• Nonlinear Acoustics
• Signal Processing

Background:

• Parametric array loudspeakers (PALs) generate directional audio via nonlinear acoustics.
• Accurate modeling of PAL sound fields is essential for personal audio zones but computationally intensive.
• Existing models struggle with complexity and nonlinear distortion reduction.

Purpose of the Study:

• To develop a computationally efficient and accurate time-domain model for audible sound fields generated by PALs.
• To simplify the modeling of sound propagation in the Westervelt far field.
• To improve the design of personal audio zones by reducing nonlinear distortion.

Main Methods:

• Proposed a differential Volterra filter (Diff-VF) model based on the Westervelt equation.
• Introduced a narrow-band (NB) approximation to decouple virtual source energy density.
• Employed temporal-spatial discretization for a simplified, input-independent model.

Main Results:

• The narrow-band Westervelt solution showed effectiveness compared to exact solutions.
• The Diff-VF model demonstrated superior prediction performance over existing Volterra filter models.
• The Diff-VF model exhibited insensitivity to input signals and reduced computational complexity.

Conclusions:

• The Diff-VF model provides an effective and simplified approach for predicting audible sound fields from PALs.
• This model facilitates the design of personal audio zones with reduced nonlinear distortion.
• The proposed method balances accuracy and computational efficiency for practical applications.