基于Gerchberg-Papoulis算法的平面近场截断误差修正方法研究

Truncation error correction method in planar near-field based on Gerchberg-Papoulis algorithm

在平面近场天线测量中,有限扫描面截断是影响测量精度的主要误差源之一,找到解决截断误差的方法是天线测量的研究重点之一.文中将平面近场天线测量中由有限区域内的场求平面波谱的过程抽象为带限函数外推的数学模型,从实际测量中的近远场变换理论出发,论证了GP(Gerchberg-Papoulis)算法应用在平面近场测量中在理论上是切实可行的.将GP算法应用在平面近场天线测量中,并分析了不同迭代次数算法的修正情况.结果表明,随着算法迭代次数的增多,可信角域外计算方向图与理论方向图差别明显减小.因此,本文的方法能够明显减小平面近场测量中截断误差的影响.除此以外,还分析了误差对算法收敛性的影响,结果表明,误差对算法修正效果影响较大.

In planar near-field antenna measurements,finite scanning plane truncation is one of the main reasons which cause measurement error,therefore,the key point is to find a solution to solve the truncation error.The process of solving the plane wave spectrum from a finite field in planar near-field measurements is abstracted as a mathematical model for the band-limited function extrapolation,and it is proved that Gerchberg-Papoulis(GP)algorithm applicated in planar near-field antenna measurements is feasible,which is based on nearfield to farfield transform theory in actual measurements.This paper introduces the GP algorithm in planar near-field antenna measurement,and the effect of algorithm with different iterations is also analyzed.The results show that the differences between deduced pattern and theoretical pattern decreased significantly out of reliable angle region with the iterative number of algorithm increase.Therefore,the method introduced in this paper can significantly reduce the truncation error.Moreover,the influences of algorithm convergence brought by errors are also analyzed,the results show that the effect of the algorithm was influenced by errors greatly.