基于近场声全息的封闭空间面板振动对声品质贡献及影响机理研究

With the development of industrial technology, the aim of noise control of the interior sound field is transformed from the reduction of the A-weighted sound pressure level into improvement of sound quality based on low noise level. However, due to the fact that the traditional method of panel acoustic contribution is only limited to the physical acoustic characteristics and can not provide the characteristics of sound quality, the relationship between the vibration behavior of the panel and the sound quality of the interior sound field is unclear. In this project, the influence mechanisms of vibrating panel on sound quality of the interior sound field in enclosure based on nearfield acoustic holography will be proposed and researched. The study was presented to reveal the distributing law of sound quality parameters and its variety law. Moreover, an efficient and accurate identification method of panel sound quality contribution to the interior sound field in enclosure was established. The key problems which will be studied in the project contain the reconstruction algorithm of the normal velocity on the surface of vibrating panel, the calculation method of physical acoustics contribution of the panel, mapping theory of the sound quality, the establishment of model algorithm for subjective evaluation system of sound quality and other key issues. The procedure of the sound quality of the interior sound field based on the near field acoustic holography is further built. On the basis of the theoretical studies, a measuring and analyzing system is developed and utilized for experimental studies finally. The results in this project are expected to provide important theoretical basis and technical support for the design and control of sound quality of the interior sound field in engineering practice.

随着工业技术的发展，封闭空间内部声场的控制目标已由传统的降低声音信号的A计权声压级转换为低噪声基础上声音品质的改善，然而传统的封闭空间面板声学贡献度分析方法只局限于物理声特征，无法提供声品质特征，因此目前对组成声腔的面板振动特性与内部声场声品质之间的联系尚不清楚。为此，本项目拟利用近场声全息方法研究封闭空间内部声场各振动面板对声品质特性的影响机理，揭示声品质参量的分布及变化规律，建立高效且精确的封闭空间内声场的面板声品质贡献度识别方法。对封闭空间结构面板振速重建算法，物理声学贡献度计算方法，内部声场的声品质映射理论以及声品质主观评价系统模型算法等关键问题展开系统研究，形成完善的基于近场声全息的封闭空间内声场声品质的计算理论和评价流程。基于以上理论，开发相应的测量分析系统，开展实验研究。本研究将为解决工程实际中封闭空间内声品质设计与控制问题提供重要的理论基础和技术支持。

封闭空间内部声场的控制目标已由传统的降低声音信号的A计权声压级转换为低噪声基础上声音品质的改善，然而传统的封闭空间面板声学贡献度分析方法只局限于物理声特征，无法提供声品质特征。为此，本项目利用近场声全息方法研究封闭空间声场振动面板对声品质的影响机理，建立封闭空间内声场的面板声品质贡献度识别方法。提出了基于Helmholtz方程最小二乘法近场声全息的结构面板声学信息的重建方法，获得了既能提供声学贡献的正负属性、又能与内部声场中的任意场点相互关联的新型面板物理声学贡献度。针对工程实际中往往只关心几个重点局部区域的声振特性而仍需全场测量的问题，提出了基于声场分离和Patch全息测量的面板表面声压和法向振速重建算法，以振动面板辐射声功率为指标的评价物理声学贡献度。以空间二维Fourier变换的近场声全息算法为基础，建立基于稀疏贝叶斯学习方法的全息数据插值模型与外推模型，有效抑制了测量孔径与测量间隔对重建精度的限制。为解决全息重建逆运算求解过程中存在的不适定性问题，提出了基于虚拟声源法的Tikhonov正则化最优参数的选择准则，确保全息重建计算能够获得准确和稳定性正则化参数。为解决Patch近场声全息中重建分辨率与测量成本的问题，建立了基于快速贝叶斯压缩感知的声场稀疏重建算法，实现以较少的测点获取较高的重建空间分辨率。将近场声全息技术和声品质客观参量计算方法相结合，提出了声品质客观参量重建算法，推导了声品质贡献度计算方法。构建了基于BP神经网络算法的封闭空间内部声场的声品质评价模型，利用遗传算法优化模型的初始权值和阈值，提高了模型预测的精度。在理论和实验研究的基础上，在LabView平台上初步研制开发了一套基于近场声全息的声品质参量测试分析系统。本项目的研究成果将为封闭空间内部声场的声品质分析评价和低噪声优化设计提供理论依据。