论文部分内容阅读
目前在化学计量学方面文献报道的含噪分析信号处理方法一般有:曲线拟合、卷积平滑、Fourier滤波、Karman滤波、匹配滤波、最小二乘和样条平滑等,并且对某些系统取得了一些有意义的结果.但这些方法具有一定的局限性和缺点,如Fourier方法,要知道全部处理信息及有用信息和噪声信息的频率分布状况,还要经过Fourier正、反变换,只能对信号作离线处理;Karman及匹配滤波器要求对处理信号的了解比Fourier方法更多,才能有较好的滤波效果;其余的滤波器,由于方法本身的局限性,使滤波效果只能达到一定的精度,或是使问题不
At present, the methods of noise analysis signal processing reported in the literature of stoichiometry generally include curve fitting, convolution smoothing, Fourier filtering, Karman filtering, matched filtering, least squares and spline smoothing, etc., and for some systems However, these methods have some limitations and shortcomings. For example, the Fourier method needs to know the frequency distribution of all information and useful information and noise information, The signal is processed offline; Karman and matched filters require more understanding of the processed signal than the Fourier method in order to have a better filtering effect; the rest of the filter, due to the limitations of the method itself, can only achieve a certain filtering effect Accuracy, or the problem is not