Sound field reproduction aims to reconstruct a desired sound field in a listening area by using multiple loudspeakers. To improve sound field reproduction performance, several loudspeaker placement methods have been proposed to select suitable loudspeaker positions. However, most conventional loudspeaker placement methods do not consider the distribution characteristic of the desired sound fields such as the distribution of directions from where the sound wave emanates, making these methods not optimal when the prior knowledge of the distribution characteristic of the desired sound fields is known. In this paper, a weighted loudspeaker placement method is proposed, which considers optimizing the objective function that depends on a set of desired sound fields with different weights. This paper formulates the optimization problem for weighted loudspeaker placement, presenting three alternative discretization methods to obtain a more practical optimization problem, after which the group greedy method is used to solve the optimization problem to find the near-optimal loudspeaker setup. To validate the proposed method, numerical experiments aiming to reconstruct the desired sound wave from a restricted range are conducted. The results indicate that the proposed method performs well in both narrowband and broadband cases, because it can use prior knowledge of the distribution characteristic of the desired sound fields.

中文翻译：

---

声场再现的加权扬声器放置方法


声场再现旨在利用多个扬声器在听音区域重建期望的声场。为了提高声场再现性能，已经提出了几种扬声器放置方法来选择合适的扬声器位置。然而，大多数传统的扬声器放置方法没有考虑期望声场的分布特征，例如声波发出的方向的分布，使得这些方法在期望声场的分布特征的先验知识不明确的情况下不是最佳的。已知。本文提出了一种加权扬声器放置方法，该方法考虑优化取决于一组具有不同权重的所需声场的目标函数。本文提出了加权扬声器放置的优化问题，提出了三种替代离散化方法以获得更实用的优化问题，然后使用群贪心法来求解优化问题以找到接近最优的扬声器设置。为了验证所提出的方法，进行了旨在从有限范围内重建所需声波的数值实验。结果表明，所提出的方法在窄带和宽带情况下都表现良好，因为它可以利用所需声场分布特征的先验知识。