摘要
针对传统的去噪方法对混沌信号除噪的盲目性,及往往仅对平稳环境或缓慢变化的噪声有效的局限性,本文提出根据信号与噪声在小波域的分布特性及信号和噪声的模极大在细尺度下收敛的横坐标点来检测信号的奇异性,以分形维树理论为依据决定阈值,得到噪声在小波域中的位置以及小波系数大小实现去噪。实验结果证明此法有效可行。
Traditional method for noise reduction of chaos signal is objectless and limited, since it is usually only effective for the noise in stationary environment or istics of signal and noise in wavelet domain, the signal changes slowly. According to the distribution character- singularity is detected by the abscissa point converged by maximum mould of signal and noise in minute scale. The threshold is determined according to the fractal di- mensions theory, the position of noise in wavelet domain and the wavelet coefficient are obtained for realizing noise reduction. Simulation result shows that this algorithm is effective.
出处
《电光与控制》
北大核心
2006年第6期60-63,共4页
Electronics Optics & Control
关键词
非平稳环境
混沌信号
离散小波变换
分形维数
non - stationary environment
chaos signal
discrete wavelet transform
fractal dimensions
