中文摘要：

随着便携通讯设备的快速发展,其音频系统的声聚焦特性逐渐受到了广泛的关注.受到尺寸的限制,便携通讯设备上的声场回放只能采用小尺度扬声器阵列来实现.声对比度控制法通过最大化亮区与暗区的声能量比值将声能量集中在亮区,这种算法在小尺度扬声器阵列声聚焦的实现上有着较好的应用前景.建立了5通道线性扬声器阵列模型,采用鲁棒型声对比度控制法将声能量聚焦于阵列端射方向的亮区.考虑到使用者对阵列不可避免的散射作用,重点分析了听音者散射对阵列声聚焦性能的影响,散射分析建立在一个刚性球散射模型和双刚性球散射模型（亮区和暗区同时存在听音者）的基础上.进一步采用人工头和刚性球建立了实验系统,通过实验验证了算法对于一个听音者和两个听音者散射的鲁棒性.

英文摘要：

With the rapid development of portable communication devices,acoustic focusing characteristics of audio system have attracted increasingly more attention.The sound field reproduction system on portable communication devices can only be realized using small-scale loudspeaker array due to the limited size of such devices.Acoustic Contrast Control（ACC）algorithm can focus the sound energy in the bright zone by maximizing the ratio of the mean sound energy in the bright zone to that in the dark zone,which makes it a competitive method to realize sound focusing control using a small number of loudspeakers.The robustness of the acoustic focusing algorithm has attracted many researchers＇ attention.Most of them concentrate on the influence of the imperfection of the loudspeakers and the variation of the acoustic transfer function.Considering that the scattering of the user is inevitable in a practical acoustic focusing system,it is meaningful to investigate the robustness to head scattering.In this paper,a 5-channel linear loudspeaker array model was established and the robust Acoustic Contrast Control algorithm was employed to concentrate the sound energy around listeners in the bright zone in the endfire direction of the linear loudspeaker ar-ray.The optimization was conducted in frequency domain and a time domain FIR filter was designed to realize the wide-band optimal control.Simulations and experiments using single-frequency signal and wide-band white noise signal respectively were carried out to realize this 5-channel linear loudspeaker array acoustic focusing audio system.The analysis throughout this paper concentrates on the head scattering effect on the array performance.Both onehead and two-head scattering models were utilized.The experimental system was constructed based on a dummy head and a rigid sphere and the directivity of the 5-channel linear loudspeaker array was measured to verify the robustness of the sound focusing algorithm when there were listeners around the array.