Multizone sound field reproduction for synthesizing desired sound fields has various applications. The modal matching-based method can obtain global modal coefficients to accurately reproduce the desired sound field. However, the transformation of the global modal coefficients into the loudspeaker weights for reproduction is susceptible to the ill-conditioned transformation matrix that caused by loudspeaker array positions and environment conditions. This paper proposes a multizone sound field reproduction method in the modal domain to obtain the loudspeaker weights directly. An optimization model is formulated with the objective of reproducing the desired sound field within the bright zone and with constraints on acoustic energy in whole region of the dark zone. The improved Lagrange algorithm can adaptively select the Lagrange multipliers for solving the loudspeaker weights, which are the same as the solutions obtained by the CVX method regarded as a benchmark and take less computational time. For the interzone sound interference in reproduction caused by the virtual sound propagation direction of the desired sound field in bright zone and the geometry positions of the multi zones, a realizability metric coefficient is developed based on the proposed model to evaluate the reproduction efficiency. The simulation results and experimental results indicate that the proposed method outperforms the current pressure matching and mode matching methods.

中文翻译：

---

模态域约束区域声能的多区域声场再现方法

用于合成所需声场的多区域声场再现具有多种应用。基于模态匹配的方法可以获得全局模态系数，以准确地再现所需的声场。然而，将全局模态系数变换为用于再现的扬声器权重容易受到由扬声器阵列位置和环境条件引起的病态变换矩阵的影响。本文提出了一种模态域多区域声场再现方法，直接获得扬声器权重。制定优化模型的目的是在亮区中再现所需的声场，并在暗区的整个区域中限制声能。改进的拉格朗日算法可以自适应地选择拉格朗日乘数来求解扬声器权重，与以CVX方法为基准的解相同，并且计算时间更少。针对亮区中所需声场的虚拟声音传播方向和多区域的几何位置所引起的再现中的区间声音干扰，基于所提出的模型开发了可实现性度量系数来评估再现效率。仿真结果和实验结果表明，该方法优于当前的压力匹配和模式匹配方法。