基于集成经验模态分解的海杂波去噪

针对实际海杂波信号非线性非平稳的特点,提出基于集成经验模态分解(EEMD)的海杂波去噪方法.利用EEMD将含有目标信号的海杂波数据分解成一系列从高频到低频的固有模态函数(IMF),通过各个IMF的自相关,分选出有用信号和噪声分量,对噪声占主导作用的IMF选用Savitzky Golay(SG)滤波方法进行消噪,将滤波后的模态分量和剩余的分量进行重构得到削噪后的信号.结合最小二乘支持向量机(LSSVM)建立混沌序列的单步预测模型,从预测误差中检测淹没在海杂波背景中的微弱信号,比较去噪前和去噪后的均方根误差,利用均方根误差评价去噪效果.实验结果表明,EEMD算法对海杂波数据去噪是有效的,去噪后所得的均方根误差0.0028比去噪前所得的均方根误差0.0119降低了一个数量级.

The Sea Clutter De-noising Based on Ensemble Empirical Mode Decomposition

In view of the nonlinear and non-stationary sea clutter signal, we put forward a de-noising method of sea clutter based on ensemble empirical mode decomposition(EEMD).By EEMD, sea clutter data containing the target signal can be decomposed into a series of intrinsic mode function.Noise component uses the Savitzky Golay filter method for de-noising.The mode components after filtering and the remaining components are reconstructed into a new signal.Combined with least square support vector machine, single-step prediction model of chaotic sequence is set up.Compare the root mean square error before and after de-noising so that we can evaluate the denoising effect from the root mean square error.The experimental results show that EEMD algorithm is effective for the de-noising of the sea clutter.By de-noising, the root mean square error can be reduced by one orders of magnitude, reaching 0.0028, while the model before de-noising can reach only 0.0119.