一种利用分布式传声器阵列的声源三维定位方法

发布时间：2018-04-28 16:19

本文选题：声源定位 + 传声器阵列　；参考：《声学学报》2017年03期

【摘要】：为了提高噪声和混响条件下分布式传声器阵列进行声源定位的性能,提出一种利用空间稀疏性和压缩感知原理的声源三维定位方法。该方法首先通过两次离散余弦变换方式提取出声音信号特征,并用该特征来构建稀疏定位模型,以便能够综合利用语音信号的短时和长时特性,同时降低模型维数;然后利用在线字典学习技术动态调整字典,克服稀疏模型与实际信号之间的失配问题,增强稀疏定位模型的鲁棒性;进而提出一种改进的平滑l_0范数稀疏重构算法来进行声源位置解算,以提高低信噪比条件下的重构精度。仿真结果表明该方法不仅可以实现多目标定位,而且具有较强的抗噪声和抗混响能力.
[Abstract]:In order to improve the performance of distributed microphone array for sound source location under noise and reverberation conditions, a 3D sound source location method based on spatial sparsity and compressed sensing principle is proposed. Firstly, the feature of sound signal is extracted by twice discrete cosine transform, and the sparse localization model is constructed with the feature, so that the short and long time characteristics of speech signal can be integrated and the dimension of the model can be reduced. Then the online dictionary learning technology is used to dynamically adjust the dictionary to overcome the mismatch between the sparse model and the actual signal and to enhance the robustness of the sparse location model. Furthermore, an improved smoothing algorithm of LS-norm sparse reconstruction is proposed to calculate the location of the sound source, so as to improve the reconstruction accuracy under the condition of low signal-to-noise ratio (SNR). Simulation results show that this method can not only achieve multi-target localization, but also has strong anti-noise and anti-reverberation ability.
【作者单位】：南京师范大学江苏省声频技术工程实验室;南京师范大学江苏省光电技术重点实验室;江苏省地理信息资源开发与利用协同创新中心;
【基金】：高等学校博士学科点专项科研基金项目(20133207120007) 国家自然科学基金项目(61405094) 江苏省气象探测与信息处理重点实验室开放基金项目(KDXS1408) 江苏省科技支撑计划-工业部分(BE2014139)资助
【分类号】：TN641;TN912.3

【参考文献】