Acoustic inversion method based on the shear flow Green's function for sound source localization in open-jet wind tunnels

While localizing sound sources within the shear flow, conventional beamforming faces limitations due to the Rayleigh criterion, restricting its resolution. Moreover, acoustic inversion method encounters challenges in establishing the relationship between source strength and acoustic quantities within the shear flow, considering the effects of convection, refraction, and reflection. This paper introduces a novel approach, acoustic inversion method based on the shear flow Green's function, to address sound source localization. The function is derived to mathematically describe the propagation of acoustic pressure and particle velocity in the shear flow, accounting for flow-acoustic effects such as convection, refraction, and reflection. Using the derived function, the transfer matrix relating the source strength to the measured acoustic pressure or particle velocity is constructed. The ℓ₁ norm constrained minimization and sparse Bayesian learning are then applied to estimate source strength distribution and achieve accurate localization. The acoustic propagation simulations and wind tunnel experiments demonstrate that the method can go beyond the Rayleigh limit, providing higher resolution than conventional beamforming techniques. And the method employs a more reasonable source-to-receiver transfer model, resulting in superior performance to other shear flow correction method. Notably, particle velocity exhibits superior localization accuracy and robustness in the wind tunnel experiments, reducing relative localization errors by up to 26.7% compared to acoustic pressure.