基于近场声全息的重建算法分析与研究

针对声全息算法种类繁多及应用场合不同需求,通过有限元仿真和数值仿真相结合,对基于傅里叶变换、统计最优和等效源3种算法进行分析,寻找声源频率、重建距离、采样间距及正则化方法对重建精度的影响,并对其计算效率进行对比。在开阔水域进行实验验证。结果表明:随着声源频率增大,重建距离增加,采样点数减少,声全息算法的重建精度逐渐降低。在低频区域,结合L-曲线正则化法的统计最优近场声全息具有最佳的声场重建效果;基于等效源法的声全息重建精度最高,但容易产生虚像;基于傅里叶变换的声全息算法受重建距离影响严重,但重建速度优异,且声源定位准确。

Analysis and Research of Reconstruction Algorithm Based on Near-field Acoustic Holography

The wide variety of acoustic holography algorithms and the requirements of different applications, the three algorithms based on Fourier transform, statistically optimized and equivalent source model are analyzed by combining finite element simulation and numerical simulation. The influence of source frequency, reconstruction distance, sampling distance and regularization methods on reconstruction accuracy are found, and their computational efficiency is compared. Experiments are carried out in open water. The results show that: with the sound source frequency and the reconstruction distance increases, the number of sampling points decreases, the reconstruction accuracy decreases gradually. In the low frequency region, the statistical optimal near-field acoustic holography combined with L-curve regularization method has the best sound field reconstruction effect; the acoustic holography reconstruction based on equivalent source method has the highest accuracy, but it is easy to produce virtual images; the acoustic holography algorithm based on Fourier transform is seriously affected by the reconstruction distance, but the reconstruction speed is excellent, and the sound source positioning is accurate.