基于神经网络的直接定位互耦与幅相差校正  

Direct Positioning Mutual Coupling and Amplitude&Phase Error Correction Based on Neural Network

作　　者：祝志聪 张莉[1] ZHU Zhicong;ZHANG Li(Information Engineering University,Zhengzhou 450001,China)

机构地区：[1]信息工程大学,河南郑州450001

出　　处：《信息工程大学学报》2022年第4期408-414,共7页Journal of Information Engineering University

摘　　要：在直接定位方法中,阵元通道不一致导致的幅相误差和实际工程中不可避免的阵元互耦效应,都会使得多重信号分类(MUSIC)定位算法性能大大降低。在短波单站定位的场景下,分别对这两种阵列误差造成的相位误差和幅度误差进行分析,证明相位特征对直接定位精度的重要性,进而建立一个增强阵列相位特征的神经网络,从而在利用协方差矩阵进行空间谱估计时,降低因畸变相位特征造成的定位误差。实验结果表明,提出的方法大大提高了阵列存在互耦和幅相差的情况下,MUSIC定位算法的精度。In the direct positioning method,the amplitude and phase error caused by the inconsistent channel gains of the array elements and the inevitable mutual coupling effects of the array elements in actual engineering will greatly reduce the performance of the multiple signal classification( MUSIC) positioning algorithm.This paper analyzes the channel amplitude and phase error and the array amplitude error and phase error caused by the mutual coupling of array elements in the scenario of shortwave single station positioning,and proves the importance of the phase feature to the direct positioning accuracy.Further,a neural network that enhances the phase feature of the array is established.Accordingly,when using the covariance matrix to estimate the spatial spectrum,the positioning error caused by the distorted phase characteristics is reduced.The experimental results show that the proposed method greatly improves the accuracy of the MUSIC positioning algorithm when the array has mutual coupling and amplitude phase error.

关 键 词：神经网络 直接定位 MUSIC定位算法 幅相误差 阵元互耦 相位特征 

分 类 号：TN918.1[电子电信—通信与信息系统]