Adaptive sparse basis compressive equivalent source method for sound field reconstruction.

Area of Science:

• Acoustics
• Signal Processing
• Computational Physics

Background:

• Compressive sensing enables sound field reconstruction using sparse representations.
• Overcomplete dictionaries combine bases for different source types.
• Existing methods struggle with selecting appropriate sparse bases, limiting reconstruction accuracy.

Purpose of the Study:

• To develop an adaptive sparse basis compressive equivalent source method.
• To improve the accuracy and robustness of sound field reconstruction for unknown sources.
• To address the limitations of fixed sparse basis selection in conventional methods.

Main Methods:

• Constructing an overcomplete dictionary by combining sparse bases.
• Introducing joint sparsity and low-rank constraints.
• Formulating reconstruction as a Bayesian optimization problem promoting sparsity and low-rank structures.

Main Results:

• The proposed adaptive method effectively selects suitable sparse bases.
• Achieved higher reconstruction accuracy compared to conventional methods, especially for sparse and combined sources.
• Demonstrated increased robustness in sound field reconstruction.

Conclusions:

• The adaptive sparse basis method enhances sound field reconstruction accuracy and robustness.
• It offers a practical solution for noise source identification without prior source knowledge.
• This approach overcomes limitations of fixed basis selection in compressive sensing applications.