基于子空间拟合的声矢量阵阵型校正算法及DOA估计

Correction Algorithm of Acoustic Vector Sensor Array and DOA Estimation Based on Subspace Fitting

将声场中的声压、振速进行联合信息处理,基于信号子空间拟合原理提出了一种声矢量阵高分辨方位估计算法。该算法在估计性能上与声矢量阵MUSIC算法接近,但在运算量方面有所减少;并针对阵元位置存在误差时文中算法性能恶化的问题,基于迭代优化原理提出了一种声矢量阵阵型校正算法。该算法依赖两个方位未知的disjoint信源,结合单矢量传感器方位估计技术实现了对阵元位置误差的有效估计。仿真结果表明,随着disjoint信源信噪比的增大和快拍数的增加,声矢量阵阵型校正算法的性能明显提高。

Considered the advantage of combined processing by using pressure and particle velocity,a method for high resolution DOA estimation based on subspace fitting（vector sensor array subspace fitting,VSF） was proposed.The proposed method offered not only a lower complexity of calculation,but also an acceptable estimation and resolution performance.Aimed at the performance depravation of the method when the array element location error exists,an array correction method based on iterative optimization was proposed.The simulation results demonstrate that the correction method combining DOA estimation technology can effectively estimate the elements＇ location error by using single vector hydrophone and two disjoint sources,and the performance of the proposed correction method can be improved with the increase of signal-noise ratio of disjoint sources.