基于多尺度自适应卡尔曼滤波的分形信号去噪  

Fractal Signal Denoising Based on Multiscale Adaptive Kalman Filtering

作　　者：李强[1] 吴文进[1] 

机构地区：[1]安庆师范学院,安徽安庆246011

出　　处：《弹箭与制导学报》2009年第5期212-214,220,共4页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：安徽省教育厅自然科学基金(KJ2007B001)资助

摘　　要：针对1/f分形信号在传输过程会受到噪声污染问题,提出了一种基于小波变换与自适应卡尔曼滤波相结合的多尺度自适应去卷积滤波方法。根据小波变换对1/f分形信号进行白化的特点,在小波域对小波系数建模,先利用最小二乘法求解模型系数,然后基于自适应卡尔曼去卷积滤波对1/f分形信号进行估计。仿真实验表明,在不同的输入信噪比与分形指数情况下,所提出的去噪方法均能取得较好的效果。A multi-scale adaptive Kalman deconvolution filtering algorithm based on wavelet transform was proposed for fractal signal restoration. After the wavelet coefficient model of 1/f fracral signal whitened with wavelet was given, the coefficients in the model were computed by the least square method, then the adaptive Kalman deconvolution filter was used to estimate 1/f fractal signal. The simulation results show that the method has excellent ability to restore fraetal signal with different input SNR and fraetal exponent.

关 键 词：小波变换 分形信号 卡尔曼滤波器 自适应滤波 

分 类 号：TN713[电子电信—电路与系统] TN911