一种核信息系统的随机信号降噪处理技术

针对一种核信息系统随机信号的特点及噪声分析,借助于Matlab信号分析工具,通过分别采用FFT频域滤波法、抑制细节系数滤波法、全局阈值滤波法和分层阈值滤波法等对随机核信号进行降噪处理及分析比较,研究结果表明,分层阈值滤波法降噪得到的结果光滑性和相似性都比较好,同时采用最小二乘法对分层阈值降噪后数字核信号进行非线性拟合,获得核信号的理想方程,提高了拟合后的核信号的测量分析精度。

De-noise processing for random signals in nuclear information systems

Various wave filter methods were employed to de-noise and analyze random signal of nuclear systems with Matlab software based on its properties.These methods included using the following filters: an FFT frequency filter,a restraining detail coefficient filter,a global threshold filter,and a hierarchical threshold filter.The test results show that the hierarchical threshold filter could lead to the smoothest and comparable results.At the same time,the least squares method was used to fit non-linearly the digital nuclear signal and the hierarchical threshold filter.The ideal nuclear signal function was thus obtained,and the measurement analysis precision of the fitted nuclear signal was improved dramatically.