摘要
声场的局部测量不能满足基于快速傅里叶变换近场声全息理论推导的前提条件,所以该方法无法实现局部声场的精确重建。统计最优近场声全息在空间域直接实现声场的重建,避免由于使用快速傅里叶变换而产生的各种误差。结合不同的正则化方法,研究了统计最优近场声全息对局部声场的重建效果,分析了重建面边缘区域以及中心区域误差对总误差的贡献。仿真与实验结果表明:统计最优近场声全息可以实现局部声场的精确重建,重建面边缘区域的误差大于中心区域的误差;正则化技术方面,基于Engl误差最小化原则的正则化参数选择法,使得Tikhonov正则化方法更为实用。
In order to reconstruct sound field, the FFT-based near-field acoustical holography ( NAH ) demands that the measurement surface must extend to a region where the sound pressure decreases to a low level. So this method is unfit for reconstructing partial sound field in which the measurement aperture size is limited either by physical necessity or as a way of reducing the measurement cost. Statistically optimal near-field acoustical holography ( SOANH) performs the plane to plane calculations directly in the spatial domain, avoiding all errors occurred in the FFT-based NAH and significantly increasing the accuracy of reconstruction of partial sound field. In the present work, combined with the different regularization methods, the SOANH is performed for reconstructing partial sound field and the errors over the central and the peripheral sections of the reconstruction area are examined separately. The simulations and experiments show that the SONAH is successful in reconstructing partial sound field and the errors over the central sections are smaller than that over the peripheral sections. The Tikhonov regularization in conjunction with the Engl's criterion is more suitable for reconstruction of practical sound field.
出处
《振动与冲击》
EI
CSCD
北大核心
2007年第12期138-140,共3页
Journal of Vibration and Shock
基金
国家自然科学基金资助项目(50335030)
国家高技术研究发展计划(863计划
2006AA04Z175)
关键词
统计最优近场声全息
局部声场
反问题
正则化
正则化参数
statistically optimal near-field acoustical holography
partial sound field
inverse problem
regularization methods
regularization parameter
