频谱校正插值法综述及在强噪声下的精度比较

插值法是离散频谱校正分析中应用最为广泛、研究最为深入、算法种类最多的一种校正方法,按照采用的谱线类型又可以分为基于幅值谱比值和基于复数谱比值的两种插值法。在实际工程应用中,抗噪声性能是评价一种算法最为重要的指标之一。在回顾国内外插值校正算法的理论研究和发展现状的基础上,简要介绍了各种插值算法的原理和特点。通过仿真考察了各算法的抗噪声性能,重点分析了插值极性错误对目前各种算法的影响。通过分析比较,指出了现有插值法的不足之处和亟待解决的问题,对插值校正算法的发展进行了展望和探讨。

A review of interpolation algorithms in discrete spectrum correction and a comparative study of their accuracy under strong noise

Interpolation DFT (discrete Fourier transform) is the most widely applied and deeply studied approach in spectrum correction methods. A lot of interpolation algorithms were proposed in past decades, which can be divided into amplitude-spectrum-based algorithms and complex-spectrum-based algorithms. Anti-noise performance is one of the most important indexes for evaluating an algorithm in the practical engineering applications. On the review of the theory and the development in recent years of interpolation methods, the principles and characteristics of different algorithms were briefly introduced. In addition, the anti-noise performance of algorithms was investigated through computer simulations. The problem of incorrect polarity estimation was particularly analyzed. At last, the deficiency and problems in the current interpolation approaches were pointed out, and the future developments and prospects of interpolation algorithms were discussed.