基于低阶近似分布源的目标方位估计  

Low Order Approximation Distributed Sources Model and Direction-of-arrival Estimation

作　　者：李强[1] 李志舜[1] 

机构地区：[1]西北工业大学航海学院,陕西西安710072

出　　处：《电声技术》2007年第12期54-60,共7页Audio Engineering

摘　　要：广义阵列流型描述的分布源模型,利用了一阶Taylor级数展开,使得模型与分布源空间能量分布形式无关,能够更加广泛地在实际中得到应用,称之为一阶近似分布源模型。研究表明,此种模型忽略高次Taylor级数项后带来的模型误差以及后续方位估计算法的性能损失,因此提出了二阶近似分布源模型,即利用二阶Taylor级数展开。新模型能够进一步减小模型误差,并且对相应的方位估计算法带来较大的性能提升。将分布源的一阶近似模型和二阶近似模型统称为低阶近似分布源模型。随后分析了低阶近似分布源模型在方位估计中的应用,提出了广义SMVDR算法。通过计算机仿真,验证了低阶近似分布源模型方位估计算法的性能,并且研究了分布源的低阶近似模型、空间频率模型和低阶Jacobi-Anger(JA)级数展开模型的模型误差。The first order Taylor series expansion is used in the distributed sources model with generalized array manifold. Thus the model has no relationship with the form of the spatial distribution of the distributed sources and has more practicability. Here this model is called one order approximation distributed sources model. Research shows that the model error generated by neglecting the high order items of Taylor series is distinct. As a result, the performance of corresponding direction-of-arrival estimation algorithms will degrade. The two orders approximation distributed sources model with two orders Taylor series expansion is proposed. The new model can reduce the model error greatly and enhance the performance of direction-of-arrival estimation. Here the one order approximation model and two orders approximation model are called low order approximation model. Then the direction-of-arrival estimation algorithms with low order approximation model are analyzed and the generalized SMVDR algorithm is proposed. Computer simulations validate the performance of these algorithms. And the model error of low order approximation distributed sources model, spatial frequency distributed sources model and Jocabi-Anger series expansion distributed sources model are discussed.

关 键 词：分布源 广义阵列流型 TAYLOR级数 方位估计 Jacobi—Anger级数 

分 类 号：TN911[电子电信—通信与信息系统]