基于自适应提升小波变换多分辨率数据融合

建立了基于小波去噪的多分辨率多传感器数据融合模型 ,引入了提升法自适应离散小波变换 ,根据最小均方 (LMS)自适应法确定伯恩斯坦 (Bernstein)提升滤波器的权系数 ,使其匹配低分辨率传感器的数据序列 ,接着使其对高分辨率采样数据的小波分解的尺度系数进行数据更新 ,实现不同分辨率数据的融合。最后对不同分辨率三传感器测量系统进行了数值仿真。实验结果表明 ,该方法可以有效地实现多分辨率多传感器数据融合 ,而且消除了噪声干扰 ,提高了系统的测量精度。

Adaptive Lifting Wavelet Transform-Based Multiresolution Multisensor Data Fusion

A multiresolution multisensor data fusion model based on wavelet denoising is developed and adaptive discrete wavelet transform based on the lifting scheme is introduced. The weighed coefficients of Bernstein lifting filter adaptively match low resolution sampling signals by LMS(least mean square) criterion. Then scale coefficients, of which high resolution sampled data are decomposed by the lifting wavelet transform, are updated, thus achieving multiresolution data fusion. Finally, numerical simulations are done on a three-sensor multiresolution measuring system, the results show that the algorithm is efficient to realize the multiresolution multisensor data fusion, reduce the noise interference and enhance the measuring precision.