An experimental study on transfer function estimation using acoustic modelling and singular value decomposition.

Area of Science:

• Acoustics
• Signal Processing
• Computational Mechanics

Background:

• Transfer functions are crucial for sound field control, linking sound sources to sound pressure.
• Modal domain methods like SHD and SVD have been proposed for transfer function estimation.
• Numerical simulations previously compared the SHD and SVD methods.

Purpose of the Study:

• To experimentally demonstrate the feasibility of the SVD method for transfer function estimation using limited measurements.
• To assess the geometric flexibility of the SVD method compared to the SHD method.
• To showcase an application of the SVD method in acoustic contrast control.

Main Methods:

• The study employed the singular value decomposition (SVD) method for acoustic model building and basis function extraction.
• Transfer functions were estimated experimentally using limited microphone measurements within a target region.
• The SVD method's performance was evaluated across various microphone placements and system configurations.

Main Results:

• The SVD method successfully estimated broadband transfer functions up to 3099 Hz within a 0.083 m radius region using only three microphones.
• Experimental results confirmed the geometric flexibility of the SVD method, outperforming the SHD method in adaptability.
• The SVD method proved effective for acoustic contrast control applications.

Conclusions:

• The SVD method is a feasible and geometrically flexible approach for estimating sound transfer functions with limited measurements.
• This technique facilitates broadband sound zone control, even with a minimal number of microphones.
• The SVD method presents a promising solution for practical acoustic control systems.