目前用于分析INS随机噪声的方法主要有Fouirer变换、功率谱密度以及Allan方差法。针对INS信号中存在的几种噪声,分别给出了它们的功率谱密度函数和Allan方差;分析了它们在时频域表达中的不同特性;最后用模拟数据和实测数据对这3种时频域分析法进行了比较和验证。结果表明:相比于Fouirer变换和功率谱密度法,Allan方差法在随机噪声分析方面具有明显的优势,不仅能够分析出信号中不同噪声的特性,还能够确定其系数的大小。
Fouirer transform, power spectral density and Allan variance are the three methods which are often used to analyze the random noises of INS. In this paper, the power spectral density functions and Allan variance functions are presented for five kinds of random noises of INS, the characteristics of which are analyzed. Two examples of simulated data and real data are given to show that compared with fouirer transform and power spectral density, the Allan variance can not only distinguish the different random noises but also get the coefficients of the noises.