A Dual Convolution Model for Grating Lobe Prediction in Directivity of Parametric Array Loudspeakers

There has been a growing interest in the evolution of steerable parametric array loudspeakers for manipulating highly directional audio beams without mechanical steering. Given this potential, understanding sound directivity is the basis for designing a successful steerable parametric array loudspeaker. Current directivity models face challenges in predicting grating lobes due to the neglect of key factors such as transducer spacing. To address this, we propose a sound field modeling framework tailored for parametric array loudspeakers using Westervelt’s formula. Within this framework, a pioneering Dual Convolutional Model is introduced to account for critical parameters of the phased array. The efficacy of the proposed model, compared with traditional convolutional approaches, has been validated through benchmark studies and experimental measurements demonstrating high accuracy in predicting grating lobes.