【作者】 张兴周； 田金超；

【Author】 ZHANG Xing-zhou, TIAN Jin-chao (School of Information and Communication Engineering, Harbin Engineering University, Harbin 150001, China)

【机构】 哈尔滨工程大学信息与通信工程学院；

【摘要】 利用局部投影算法中局部协方差矩阵存在特征值突变的现象,提出了改进的噪声抑制方法,解决了原算法中噪声空间和信号空间特征向量难以区分的问题,并运用到Henon映射混沌信号的噪声抑制中。通过仿真研究证实,在低信噪比条件下,改进后的算法与原算法相比,降低了绝对误差,提高了噪声抑制的有效性。更多还原

【Abstract】 By means of the eigenvalues leap phenomenon that exists in local covariance matrix in local-projection algorithm, an improved method for restraint noise in contaminated chaotic signals was proposed. This method overcame the difficulty that it’s hardly to distinguish the eigenvectors between noise space and signal space in the original, and was applied to the Henon-map chaotic signal that contained noise. Through simulation research, under the mild signal-to-noise ratios conditions, the improved algorithm is known to reduce the absolute error and enhance capability of noise restraint compared with the original.更多还原