麦克风阵列论文：基于麦克风阵列的语音增强系统设计.docVIP

麦克风阵列论文：基于麦克风阵列的语音增强系统设计 【中文摘要】语音增强的是去除语音信息中的噪声成分，它是语音信号处理领域中主要的研究内容之一。然而，由于声源环境往往较为复杂，一般单一麦克风很难清晰的从环境中获取期望语音信号并进行增强处理。因此，需要进行有效的噪声抑制，以增强语音信号的质量。本文对基于麦克风阵列的语音增强算法进行研究，重点分析了时延估计和波束形成两种技术。在时延估计方面，研究了相位变换加权的广义互相关时延估计算法性能，介绍了将广义互相关算法由双麦克风扩展应用至多麦克风的优化方法，给出了算法在0dB~10dB的信噪比下的性能分析。在波束形成方面，研究了固定波束形成和自适应波束形成的算法性能，给出了不同麦克风阵列参数条件下的波束模式性能分析；并分析了在MVDR自适应波束形成算法中进行前置卡尔曼滤波的可行性，对卡尔曼滤波进行了仿真测试。最后本文给出了完整的麦克风阵列系统设计，并对该系统的硬件和软件设计进行了论述，基于麦克风阵列的硬件电路具有功耗低、扩展性强等特点，可以较好的为麦克风阵列语音增强算法提供多通道语音信号。 【英文摘要】The main aim of speech enhancement is to pick up speechinformation from the speech signals with noise, in order to obtain highquality speech. But due to the existence of the noise diversity andenvironment reverberation, the speech quality received by microphone isnot so good, which affects the speech intelligibility and the speechprocessing performance. So the effective noise suppression is necessaryto improve the speech signials quality.The work in this paper focuses on two key techniques based onmicrophone array: time delay estimation and beamforming. In the aspectof time delay estimation, the performance of phase transform weightedgeneralized cross correlation method is analyzed. Generalized crosscorrelation algorithm by dual microphones extended up to the multimicrophones is analysed. The performance analysis is given under0dB~10dB SNR. In the aspect of beamforming, an optimization method basedon Kalman-filtering and MVDR adaptive beamforming is analysed.Simulation shows the validity of this method. Finally, this paper presentsa viable microphone array speech enhancement system design. Theimplementation of the software and hardware of the system is dealt with in detail. Microphone array hardware circuit has low power consumptionand expansibility, which can provide multichannel speech signals. 【关键词】麦克风阵列 语音增强 时延估计 波束形成 数字信号处理器 【英文关键词】Microphone Array Speech Enhancement Timedelay estimation Beamforming DSP 【目录】基于麦克风阵列的语音增强系统设计 摘要 4-5 ABSTRACT 5-6 目录 7-10 Contents 10-13 第一章