非自由场内声场定量重现的L曲线滤波方法  

L-curve-filtering method for sound field reproduction in a general room

作　　者：彭博[1] 郑四发[1] 廖祥凝 连小珉[1] 

机构地区：[1]清华大学汽车安全与节能国家重点实验室,北京100084

出　　处：《振动与冲击》2015年第15期55-59,共5页Journal of Vibration and Shock

基　　金：国家自然科学资金资助项目(51275262)

摘　　要：研究了非自由场一般房间内利用扬声器阵列进行声场重现的方法。建立了房间内声场重现的系统模型,利用正则化最小二乘法和L曲线法获得了实现声场内多个控制点定量重现的驱动信号频域特性,然后采用L曲线滤波器设计得到了阵列扬声器时域驱动信号。实际汽车噪声的重现结果表明:L曲线滤波方法能够实现扬声器驱动信号功率与声场重现误差之间的平衡,可以在非自由场内实现稳态声场的定量重现。L-curve-filtering method for sound field reproduction in a general room with a loudspeaker array was studied. Firstly,the system model used for the sound field reproduction in a room was established. Then,the regularized least-square method and L-curve method were used to obtain the filtering properties in frequency domain of drive signals of speakers,they reproduced the sound pressures of multiple controlled points quantitatively. Finally,a L-curve-filter was designed to obtain the time domain drive signals of the loudspeaker array. The test results to reproduce a vehicle noise showed that L-curve-filtering method can balance the power of drive signals and the sound field reproduction error to realize the sound field reproduction in a general room.

关 键 词：声场重现 滤波方法 L曲线方法 

分 类 号：TN912.27[电子电信—通信与信息系统]