分布式麦克风阵列多目标分区优化校准方法
- 分享到:
摘要：
为提升语音分布式阵列的噪声抑制和有效语音信息的获取能力,针对阵列麦克风测量信号在时间和相位上的不同步及噪声污染问题，提出一种分区多目标优化的校准方法.首先，对麦克风所在位置分区.以麦克风初始位置为圆心,人射波波长为初始半径,以麦克风信噪比为权重优化半径自适应分区,其次,以每个麦克风的分区区域为搜索范围,设置阵列麦克风最大基线、阵列增益、信号幅相误差、输出波束的主瓣宽度和旁瓣抑制比为优化目标,构建阵列校准多目标优化函数.由鲁洛孔径提供最大基线优化方向.最后,引入多目标粒子群优化算法，对麦克风位置寻优.进行初始分布式阵列校准.经数值仿真和实验对现有方法进行性能对比,验证了该方法的有效性。
To enhance the noise suppression and the effective speech acquisition capabilities of distributed microphone arrays, a partitioned multi-objective optimization calibration method is proposed to address issues of signal asynchrony in timing and phase, as well as noise contamination, First, the microphones are partitioned into regions based on their initial positions,with the initial radius determined by the incident wave's wavelength and adaptively optimized using the signal-to-noise ratio of each microphone as the weight. Second, the partitioned area of each microphone is used as the search range to optimize objec-tives such as maximum baseline, array gain,signal amplitude and phase error, main lobe width, and side lobe suppression ratio, forming a multi-objective optimization function for array calibration. The Reuleaux aperture is introduced to guide the opti-mization of the maximum baseline direefion, Finally, a mul-objective partide swarm optrmization algorthm is employed to op-timife mtertophone posffons for initial caibratfon of the distrfibtted aray. Numerfeal simtlafons and espertmental comparisons validate the effectiveness of the proposed method.
作者:
刘志红,刘景轲,冯志鹏,赵耀,王智勇,张海蕊
Liu Zhihong,Liu Jingke,Feng Zhipeng,Zhao Yao,Wang Zhiyong,Zhang Hairui
机构地区：
青岛理工大学工业流体节能与污染控制教育部重点实验室；机械与汽车工程学院
引用本文：
刘志红,刘景轲,冯志鹏等.分布式麦克风阵列多目标分区优化校准方法[J].河南师范大学学报(自然科学版),2026,54(1): 107-113.(Liu Zhihong, Liu Jingke,Feng Zhipeng.et al.Multi-obiective partition optimization calibration method for distributed microphone arrays[J].Journal of Henan Normal University (Natural Science Edition),2026,54(1):107-113.DOI:10.16366/j.cnki.1000-2367.2024.10.15.0003.)
基金：
国家自然科学基金；山东省自然科学基金
关键词：
分布式阵列;阵列校准;性能校准;多目标粒子群算法;分区优化
distributed array; array calibration; performance calibration; mult-objective particle swarm optimization;partition optimization
分类号：
TB52
