自适应降噪麦克风壳体声学有限元分析与设计

Housing design for adaptive noise cancellation microphones based on finite element analysis

基于DSP技术的自适应主动降噪麦克风是近年来适应超高噪声下通讯应用而发展起来的麦克风技术。该技术依靠两个独立的无指向型传声元件实现降噪。这两个传声元件通常背靠背地封装在一起。根据自适应滤波理论,系统的降噪性能与这两个元件间对应背景噪声的空间相干函数及对应近场语音的传递函数有关。这两个函数又与麦克风封装壳体的几何设计直接相关。为优化降噪性能,建立了一个以有限元声学仿真为基础的,考虑声场特性及麦克风壳体几何形状的主动降噪性能的模型。依据此模型,提出了新型附加反射板的壳体设计方案。仿真和实验结果显示,相对传统壳体设计,有限元优化后的新方案可以提高麦克风降噪性能大于15dB,提高输出信噪比大于10dB。

With the progress of the DSP technology, the adaptive noise cancellation microphone （ANCM） has been developed to meet the communication requirements in extremely noisy environments. Typically, an ANCM requires two independent onmi-directional microphone elements packaged together back-to-back. According to the adaptive filter theory, the noise cancellation performance is proportional to the spatial coherence function of the background noise field sensed by the two omni-directional microphone elements. On the other hand, the final signal-to-noise-ratio of the ANCM is also related to the transfer function between the two omni-directional microphone elements with respect to the near-field source. It is known that both the coherence function and the transfer function are influenced by the housing design of the ANCM. In this paper, a finite element （FE） model is firstly developed to simulate the performance of the ANCM. The effectiveness of this model is then demonstrated experimentally. Finally, using the FE model, a novel housing design with a reflective shell is proposed to significantly improve the ANCM performance. Both the analytical and the experimental data show that, compared with the traditional housing design, a proper reflective shell design can improve the noise cancellation performance by at least 15 dB and raise the signal-to-noise ratio in the final output by 10 dB, respectively.