• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2018, Vol. 54 ›› Issue (4): 238-244.doi: 10.3901/JME.2018.04.238

• 振动与噪声 • 上一篇    下一篇

扫码分享

基于声压贡献的球面阵波束形成声源识别滤波求和算法

褚志刚1,2, 陈涛1,2, 赵书艺1,2, 肖新标3   

  1. 1. 重庆大学机械传动国家重点实验室 重庆 400044;
    2. 重庆大学汽车工程学院 重庆 400044;
    3. 西南交通大学牵引动力国家重点实验室 成都 610031
  • 收稿日期:2017-04-05 修回日期:2017-12-18 发布日期:2018-02-20
  • 通讯作者: 褚志刚(通信作者),男,1978年出生,博士,副教授,博士研究生导师。主要研究方向为车辆系统动力学与控制、车辆振动噪声控制、计算机辅助测试理论与技术。E-mail:zgchu@cqu.edu.cn
  • 作者简介:陈涛,男,1991年出生,硕士研究生。主要研究方向为传声器阵列信号处理技术及其在车辆噪声测量中的应用。E-mail:t.chen@cqu.edu.cn
  • 基金资助:
    国家自然科学基金(11774040,11404368)、中央高校基础研究基金(106112017CDJQJ338810,CDJX22016003)和重庆市重大应用及发展计划(cstc2015yykfc60003)资助项目

Filter-and-sum Beamforming Sound Source Identification Algorithm for Spherical Microphone Arrays Based on Pressure Contrition

CHU Zhigang1,2, CHEN Tao1,2, ZHAO Shuyi1,2, XIAO Xinbiao3   

  1. 1. State Key Laboratory of Mechanical Transmission, Chongqing University, Chongqing 400044;
    2. College of Atomotive Engineering, Chongqing University, Chongqing 400044;
    3. Traction Power State Key Laboratory, Southwest Jiaotong University, Chengdu 610031
  • Received:2017-04-05 Revised:2017-12-18 Published:2018-02-20

摘要: 运用实心球传声器阵列波束形成技术识别声源时,滤波求和算法提高了传统声压球谐函数角度分解算法的旁瓣抑制性能,但其采用声源强度作为输出,无法直接反映声源对目标接收者的声学贡献,且输出结果是否准确依赖于聚焦距离是否等于源到阵列中心的真实距离。针对此问题,以最小化最大旁瓣为目标,以声压贡献输出的主瓣峰值无畸变为约束条件,构建二阶锥规划优化模型,采用CVX凸优化求解器进行求解获取滤波参数,最终建立以声压贡献为输出的球面阵波束形成声源识别滤波求和算法。仿真及试验结果均表明,该算法实现了声源对目标接收者的声压贡献量化,且量化准确度几乎不受聚焦距离的影响,同时具有优秀的旁瓣抑制性能和良好的空间分辨率。

关键词: 波束形成, 球阵列, 声压贡献, 声源识别

Abstract: While utilizing solid spherical microphone arrays beamforming technology for sound source identification, the filter-and-sum algorithm improves the performance of side lobe attenuation relative to conventional spherical harmonics angularly resolved pressure algorithm, but it uses the source strength as output, which cannot reflect acoustic contribution from source to the target receiver, and the accuracy is relied on whether the focusing distance is equal to the true distance from the source to the center of the arrays. Aiming at this problem, a second order cone programming optimum filter model is constructed, taking the minimum of the maximum sidelobe as the objective and the mainlobe peak of sound pressure contribution without distortion as the constraint. The CVX convex optimization solver is used to obtain the filter parameters and spherical microphone array beamforming filter-and-sum algorithm which outputs pressure contribution is proposed finally. Both simulation and experiment show that the algorithm can quantify the acoustic contribution by pressure contribution from source to target, and the accuracy is almost independent of the focusing distance. Furthermore, it has excellent sidelobe suppression performance and good spatial resolution.

Key words: beamforming, pressure contribution, sound source identification, spherical microphone arrays

中图分类号: