论文部分内容阅读
本文提出,在快速付里叶变换中应用窗函数修正时,不但要考虑窗函数本身渗漏作用的大小,而且要考虑经过修正后的“有效”信息量的因素,即要保持原有信号的最大信息量和主要信息量的效果。这对瞬态信号和变频率的非平稳随机信号分析尤为重要。文内提出了Y-1~#和Y-2~#两个新数据窗,通过对实测瞬态信号的应用,并与汉宁、哈明、余弦矩形、矩形和无窗函数作用的最大熵法(MEM)等进行了对比,取得了良好的结果。文内证明了著名的汉宁和哈明窗在分析瞬态信号时效果是不稳定的,提出了窗函数前沿陡度对渗漏、旁瓣影响不大,渗漏、旁瓣主要由窗形状和其后沿陡度决定的特性,同时给出了计算应用实例。
This paper proposes that in the application of window function correction in the fast Fourier transform, not only the size of the leakage function of the window function itself should be considered, but also the factor of the “effective” amount of information after correction, that is, the original signal should be maintained. The effect of the maximum amount of information and the amount of major information. This is especially important for the analysis of transient signals and variable-frequency non-stationary random signals. Two new data windows, Y-1~# and Y-2~#, have been proposed in the article, and the maximum entropy of the interaction with Hanning, Hamming, cosine rectangles, rectangles, and windowless functions has been obtained through the application of measured transient signals. The MEM method was compared and good results were obtained. The article proves that the famous Hanning and Hamming windows are unstable when analyzing transient signals. It is proposed that the steepness of the front of the window function has little effect on the leakage and sidelobes, and the leakage and side lobes are mainly shaped by windows. With the characteristics determined by the steepness behind, along with the calculation application examples.