基于时延估计的声源定位算法实验研究

摘要

In general, because the microphone array can get more information of the acoustic source, more attention is dedicated on acoustic source location. When the acoustic source near the ends of array, the uniform-linear-array’s resolution is gradually reduced. It is unreasonable that the signal incident angle is different, but the reference range is the same. By analyzing the characteristics of uniform linear array, the time delay is been used to filter the results. The experimental results verify the reasonableness of method. Because the experimental array number is less and the length of array is short, the results also have a certain reference value for the practical application.%麦克风阵列因其相对于单麦克风能够获取更多声源信息，在声源定位的应用上正日益得到人们的极大关注。均匀线阵中，针对信号入射方向靠近阵列的两端阵列分辨力将逐渐降低，利用基于时延估计的声源定位算法时，对不同的信号入射角度以相同的角度误差作为结果筛选的依据的做法，显然不合理。通过分析均匀线阵的特点，当信号以不同角度入射时，以时延点数误差作为不同角度估计结果的筛选门限，并通过实测数据实验，验证了方法的合理性。由于实验使用阵元数少且阵列长度较短，因此实验结果对阵列信号处理的实际工程应用具有一定的参考价值。