【摘要】 传统的声源定位技术分为时延估计和空间定位两步，本文将几种经典的时延估计和空间定位方法加以整合与改进，提高了系统的实时性与抗干扰能力，创新性地提出并实现了根据声源位置变化而自适应调整麦克风阵列形态的定位系统，在不降低实时性的同时，有效地降低了定位误差。本文从定位精度、实时性和抗干扰能力三个方面对系统性能进行评估，实验证明：理想条件下，该声音定位系统的不确定度在4cm以内；对噪声和混响有较好的抵抗能力，在恶劣环境下不确定度不超过6cm;单次定位耗时0.5s,可以对运动声源或脉冲声源进行定位，有较强的实时性，可以适用于大部分应用场景。更多还原

【Abstract】 The traditional sound source location technology is divided into two steps: time delay estimation and spatial location. This paper integrates and improves several classical time delay estimation and spatial location methods to improve the real-time performance and anti-interference ability of the system, innovatively proposes and implements a location system that adaptively adjusts the shape of the microphone array according to the change of sound source position, and effectively reduces the location error without reducing the real-time performance. This paper evaluates the system performance from three aspects: positioning accuracy, real-time performance, and anti-interference ability. Experiments have shown that under ideal conditions, the uncertainty of the sound positioning system is within 4cm; Has good resistance to noise and reverberation, with an uncertainty of no more than 6cm in harsh environments; A single localization takes 0.5 seconds and can locate moving or pulsed sound sources with strong real-time performance, making it suitable for most application scenarios.更多还原