嵌入式平面麦克风阵列TDOA校正方法

doi:10.3969/j.issn.0372-2112.2019.10.027

PDF(2191 KB)

电子学报 ›› 2019, Vol. 47 ›› Issue (10) : 2235-2240. DOI: 10.3969/j.issn.0372-2112.2019.10.027

科研通信

嵌入式平面麦克风阵列TDOA校正方法

赵兆, 王旭, 许志勇

作者信息 +

Correction of TDOA for Embedded Planar Microphone Arrays

ZHAO Zhao, WANG Xu, XU Zhi-yong

Author information +

文章历史 +

摘要

针对嵌入式安装麦克风阵列因壳体遮挡部分阵元而导致的实际波达时间差（TDOA）偏离直达波理论模型问题，基于实际声探测系统最常用的对称凸多边形平面阵型构造线性几何约束并结合秩2代数约束，提出了一种校正实测TDOA矩阵的凸优化闭式解，只要阵列结构中存在至少一对等长平行线，就能获得现有TDOA校正方法所没有的壳体遮挡衍射效应抑制能力，同时还能抑制随机误差和异常值的不利影响，实现复杂度低，更适用于日益普及的小型实时全向声探测系统应用.仿真结果验证了该方法的有效性.

Abstract

In order to diminish the deviation of far-field time-difference-of-arrival (TDOA) measurements from the ideal line-of-sight (LOS) model due to some microphones shadowed by the system shell of an embedded planar microphone array, which is not considered and cannot be resolved by existing TDOA correction studies, a novel convex optimization method was proposed for correcting the measured TDOA matrix. A closed-form solution was obtained by integrating the rank-2 algebraic constraint with a set of linear geometric constraints corresponding to the equilong parallel lines (EPLs) that are easily achieved in practical acoustic detection systems using approximately symmetric convex polygons as preferable array shapes.As long as the array shape contains at least one pair of linearly independent EPLs, the proposed method gains the capability of effectively suppressing the diffraction induced deviations in measured TDOAs relevant to non-line-of-sight (NLOS) array elements. Meanwhile, the adverse effects of measurement noise and TDOA outliers can be also mitigated. Considering both the above capabilities and the low computational complexity, the proposed method is more suitable for ever-growing applications of small-scale, real-time acoustic detection systems. Numerical simulations verified its effectiveness.

导出引用

赵兆, 王旭, 许志勇. 嵌入式平面麦克风阵列TDOA校正方法[J]. 电子学报, 2019, 47(10): 2235-2240. https://doi.org/10.3969/j.issn.0372-2112.2019.10.027

ZHAO Zhao, WANG Xu, XU Zhi-yong. Correction of TDOA for Embedded Planar Microphone Arrays[J]. Acta Electronica Sinica, 2019, 47(10): 2235-2240. https://doi.org/10.3969/j.issn.0372-2112.2019.10.027

中图分类号： TN971

参考文献

[1] Lombard A,Zheng Y H,Buchner H,et al.TDOA estimation for multiple sound sources in noisy and reverberant environments using broadband independent component analysis[J].IEEE Transactions on Audio,Speech,and Language Processing,2011,19(6):1490-1503.
[2] Blandin C,Ozerov A,Vincent E.Multi-source TDOA estimation in reverberant audio using angular spectra and clustering[J].Signal Processing,2012,92(8):1950-1960.
[3] 闫青丽,陈建峰.分布式声源定位系统节点最优布局方法及性能研究[J].电子学报,2018,46(5):1186-1193. Yan Qing-li,Chen Jian-feng.Node placement optimization for distributed acoustic source localization system and performance study[J].Acta Electronica Sinica,2018,46(5):1186-1193.(in Chinese)
[4] 许志勇,赵兆.平面阵声源方位角估计扰动敏感性分析[J].西安电子科技大学学报,2017,44(4):105-110. Xu Zhi-yong,Zhao Zhao.Perturbation sensitivity analysis on azimuth estimation of acoustic source for planar microphone array[J].Journal of Xidian University,2017,44(4):105-110.(in Chinese)
[5] Kim U H,Nakadai K,Okuno H G.Improved sound source localization in horizontal plane for binaural robot audition[J].Applied Intelligence,2015,42(1):63-74.
[6] Zhong X,Sun L,Yost W.Active binaural localization of multiple sound sources[J].Robotics and Autonomous Systems,2016,85:83-92.
[7] Scheuing J,Yang B.Disambiguation of TDOA estimates in multi-path multi-source environments (DATEMM)[A].ICASSP 2006[C].Toulouse,France:IEEE,2006.837-840.
[8] Scheuing J,Yang B.Efficient synthesis of approximately consistent graphs for acoustic multi-source localization[A].ICASSP 2007[C].Honolulu,USA:IEEE,2007.501-504.
[9] Scheuing J,Yang B.Disambiguation of TDOA estimation for multiple sources in reverberant environments[J].IEEE Transactions on Audio,Speech,and Language Processing,2008,16(8):1479-1489.
[10] Le T-K,Ono N.Refinement of time-difference-of-arrival measurements via rank properties in two-dimensional space[A].Proceedings of the 25th European Signal Processing Conference (EUSIPCO)[C].Kos,Greece:EURASIP,2017.1971-1975.
[11] Le T-K,Ho K C,Le T-H.Rank properties for matrices constructed from time differences of arrival[J].IEEE Transactions on Signal Processing,2018,66(13):3491-3503.
[12] Velasco J,Pizarro D,Macias-Guarasa J,et al.TDOA matrices:Algebraic properties and their application to robust denoising with missing data[J].IEEE Transactions on Signal Processing,2016,64(20):5242-5254.
[13] 余光正,刘昱,谢菠荪.近场头相关传输函数的多声源快速测量系统设计与验证[J].声学学报,2017,42(3):348-360. Yu Guang-zheng,Liu Yu,Xie Bo-sun.Design and validation on a multiple sound source fast measurement system of near-field head-related transfer functions[J].Acta Acoustica,2017,42(3):348-360.(in Chinese)
[14] 黄婉秋,曾向阳,王蕾.基于多维生理参数的头相关传递函数个人化方法[J].西北工业大学学报,2018,36(2):281-286. Huang Wan-qiu,Zeng Xiang-yang,Wang Lei.Personalization method for HRTF based on multi-dimensional physiological parameters[J].Journal of Northwestern Polytechnical University,2018,36(2):281-286.(in Chinese)
[15] Alameda-Pineda X,Horaud R.A geometric approach to sound source localization from time-delay estimates[J].IEEE Transactions on Audio,Speech,and Language Processing,2014,22(6):1082-1095.
[16] 张贤达.矩阵分析与应用[M].第2版.北京:清华大学出版社,2013:95,106. Zhang Xian-da.Matrix Analysis and Applications (2nd Edition)[M].Beijing:Tsinghua University Press,2013.95,106.(in Chinese)