瞬态声品质时域传递路径方法与应用研究

全球化竞争日趋激烈和消费者对汽车舒适性要求越来越高,迫使各汽车公司加快了NVH开发进程,汽车声音的控制逐渐进入声品质控制阶段。创新性提出一种基于时域传递路径分析的瞬态声品质分析方法和流程。采用考虑奇异值截断的去卷积滤波器方法建立时域去卷积网络。构建了车内瞬态噪声合成模型,并在时频域上分解和分析了发动机的结构声贡献和空气声贡献。通过视听比较合成噪声和测量噪声,以评审团主观评价打分的形式来验证模型的准确性。进一步对合成噪声进行主观声品质评价,将车内噪声合成模型延伸至虚拟车内声品质预测模型。基于该模型,找到声品质贡献较大的路径,并且通过虚拟修改各路径传递函数值,来优化车内声品质,为制定车内声品质改善措施提供指导依据。

Analysis of transient sound quality in time domain transfer path method and its application

The increasingly fierce global automobile market competition and consumers’ higher and higher requirements for automobile comfort have forced automobile companies to speed up the development process of NVH, and automobile sound control has gradually entered the stage of sound quality control. In this paper, a transient sound quality analysis method and process based on time-domain transfer path analysis was proposed. Firstly, the time-domain deconvolution network was established by using the deconvolution filter method considering singular value truncation. Then the interior transient noise synthesis model was constructed, and the structural-borne acoustic contribution and air-borne acoustic contribution of the engine were decomposed and analyzed in time-frequency domain. Finally, through the audio-visual comparison of synthetic noise and measurement noise, in the form of subjective evaluation of the jury to verify the accuracy of the model. Furthermore, the subjective sound quality of the synthetic noise was evaluated, and the interior noise synthesis model was extended to the virtual interior sound quality prediction model. Based on this model, the paths with bigger contribution to sound quality were found, and the transfer function values of each path were virtually modified to optimize the sound quality in the vehicle, so as to provide guidance for improving the sound quality in the vehicle.